iv/linux
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
09134c5322
linux/drivers/spi/spi-mtk-nor.c

937 lines
24 KiB
C
Raw Normal View History

spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
// SPDX-License-Identifier: GPL-2.0
//
// Mediatek SPI NOR controller driver
//
// Copyright (C) 2020 Chuanhong Guo <gch981213@gmail.com>
#include <linux/bits.h>
#include <linux/clk.h>
#include <linux/completion.h>
#include <linux/dma-mapping.h>
#include <linux/interrupt.h>
#include <linux/io.h>
#include <linux/iopoll.h>
#include <linux/kernel.h>
#include <linux/module.h>
#include <linux/of_device.h>
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
#include <linux/pm_runtime.h>
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
#include <linux/spi/spi.h>
#include <linux/spi/spi-mem.h>
#include <linux/string.h>
#define DRIVER_NAME "mtk-spi-nor"
#define MTK_NOR_REG_CMD 0x00
#define MTK_NOR_CMD_WRITE BIT(4)
#define MTK_NOR_CMD_PROGRAM BIT(2)
#define MTK_NOR_CMD_READ BIT(0)
#define MTK_NOR_CMD_MASK GENMASK(5, 0)
#define MTK_NOR_REG_PRG_CNT 0x04
spi: spi-mtk-nor: make use of full capability of prg mode "program" mode on this controller can trigger up to 56 bits of data shifting. During the operation, data in PRGDATA[0-5] will be shifted out from MOSI, and data from MISO will be continuously filling SHREG[0-9]. Currently this mode is used to implement transfer_one_message for 6-byte full-duplex transfer, but it can execute a transfer for up-to 7 bytes as long as the last byte is read only. transfer_one_message is expected to perform full-duplex transfer, instead of transfer with specific format. mtk_nor_spi_mem_prg is added here to use this extra byte. Newer version of this controller can trigger longer data shifting with shift bytes more than PRGDATA_MAX + SHREG_MAX. This patch is implemented with that in mind and it checks against both SHREG_MAX and PRG_CNT_MAX for future support of new controllers. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-2-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:28 +03:00
#define MTK_NOR_PRG_CNT_MAX 56
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
#define MTK_NOR_REG_RDATA 0x0c
#define MTK_NOR_REG_RADR0 0x10
#define MTK_NOR_REG_RADR(n) (MTK_NOR_REG_RADR0 + 4 * (n))
#define MTK_NOR_REG_RADR3 0xc8
#define MTK_NOR_REG_WDATA 0x1c
#define MTK_NOR_REG_PRGDATA0 0x20
#define MTK_NOR_REG_PRGDATA(n) (MTK_NOR_REG_PRGDATA0 + 4 * (n))
#define MTK_NOR_REG_PRGDATA_MAX 5
#define MTK_NOR_REG_SHIFT0 0x38
#define MTK_NOR_REG_SHIFT(n) (MTK_NOR_REG_SHIFT0 + 4 * (n))
#define MTK_NOR_REG_SHIFT_MAX 9
#define MTK_NOR_REG_CFG1 0x60
#define MTK_NOR_FAST_READ BIT(0)
#define MTK_NOR_REG_CFG2 0x64
#define MTK_NOR_WR_CUSTOM_OP_EN BIT(4)
#define MTK_NOR_WR_BUF_EN BIT(0)
#define MTK_NOR_REG_PP_DATA 0x98
#define MTK_NOR_REG_IRQ_STAT 0xa8
#define MTK_NOR_REG_IRQ_EN 0xac
#define MTK_NOR_IRQ_DMA BIT(7)
#define MTK_NOR_IRQ_MASK GENMASK(7, 0)
#define MTK_NOR_REG_CFG3 0xb4
#define MTK_NOR_DISABLE_WREN BIT(7)
#define MTK_NOR_DISABLE_SR_POLL BIT(5)
#define MTK_NOR_REG_WP 0xc4
#define MTK_NOR_ENABLE_SF_CMD 0x30
#define MTK_NOR_REG_BUSCFG 0xcc
#define MTK_NOR_4B_ADDR BIT(4)
#define MTK_NOR_QUAD_ADDR BIT(3)
#define MTK_NOR_QUAD_READ BIT(2)
#define MTK_NOR_DUAL_ADDR BIT(1)
#define MTK_NOR_DUAL_READ BIT(0)
#define MTK_NOR_BUS_MODE_MASK GENMASK(4, 0)
#define MTK_NOR_REG_DMA_CTL 0x718
#define MTK_NOR_DMA_START BIT(0)
#define MTK_NOR_REG_DMA_FADR 0x71c
#define MTK_NOR_REG_DMA_DADR 0x720
#define MTK_NOR_REG_DMA_END_DADR 0x724
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
#define MTK_NOR_REG_DMA_DADR_HB 0x738
#define MTK_NOR_REG_DMA_END_DADR_HB 0x73c
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
#define MTK_NOR_PRG_MAX_SIZE 6
// Reading DMA src/dst addresses have to be 16-byte aligned
#define MTK_NOR_DMA_ALIGN 16
#define MTK_NOR_DMA_ALIGN_MASK (MTK_NOR_DMA_ALIGN - 1)
// and we allocate a bounce buffer if destination address isn't aligned.
#define MTK_NOR_BOUNCE_BUF_SIZE PAGE_SIZE
// Buffered page program can do one 128-byte transfer
#define MTK_NOR_PP_SIZE 128
spi: spi-mtk-nor: fix timeout calculation overflow CLK_TO_US macro is used to calculate potential transfer time for various timeout handling. However it overflows on transfer bigger than 512 bytes because it first did (len * 8 * 1000000). This controller typically operates at 45MHz. This patch did 2 things: 1. calculate clock / 1000000 first 2. add a 4M transfer size cap so that the final timeout in DMA reading doesn't overflow Fixes: 881d1ee9fe81f ("spi: add support for mediatek spi-nor controller") Cc: <stable@vger.kernel.org> Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200922114905.2942859-1-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-22 14:49:02 +03:00
#define CLK_TO_US(sp, clkcnt) DIV_ROUND_UP(clkcnt, sp->spi_freq / 1000000)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
struct mtk_nor {
struct spi_controller *ctlr;
struct device *dev;
void __iomem *base;
u8 *buffer;
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
dma_addr_t buffer_dma;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
struct clk *spi_clk;
struct clk *ctlr_clk;
spi: spi-mtk-nor: add axi clock control for MT8192 spi-nor MT8192 spi-nor is an independent sub system, we need extra control axi bus clock for it. Add support for the additional axi clock to allow it to be configured appropriately. Signed-off-by: bayi cheng <bayi.cheng@mediatek.com> Tested-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/1605084902-13151-2-git-send-email-bayi.cheng@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-11-11 11:55:02 +03:00
struct clk *axi_clk;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
unsigned int spi_freq;
bool wbuf_en;
bool has_irq;
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
bool high_dma;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
struct completion op_done;
};
static inline void mtk_nor_rmw(struct mtk_nor *sp, u32 reg, u32 set, u32 clr)
{
u32 val = readl(sp->base + reg);
val &= ~clr;
val |= set;
writel(val, sp->base + reg);
}
static inline int mtk_nor_cmd_exec(struct mtk_nor *sp, u32 cmd, ulong clk)
{
ulong delay = CLK_TO_US(sp, clk);
u32 reg;
int ret;
writel(cmd, sp->base + MTK_NOR_REG_CMD);
ret = readl_poll_timeout(sp->base + MTK_NOR_REG_CMD, reg, !(reg & cmd),
delay / 3, (delay + 1) * 200);
if (ret < 0)
dev_err(sp->dev, "command %u timeout.\n", cmd);
return ret;
}
static void mtk_nor_set_addr(struct mtk_nor *sp, const struct spi_mem_op *op)
{
u32 addr = op->addr.val;
int i;
for (i = 0; i < 3; i++) {
writeb(addr & 0xff, sp->base + MTK_NOR_REG_RADR(i));
addr >>= 8;
}
if (op->addr.nbytes == 4) {
writeb(addr & 0xff, sp->base + MTK_NOR_REG_RADR3);
mtk_nor_rmw(sp, MTK_NOR_REG_BUSCFG, MTK_NOR_4B_ADDR, 0);
} else {
mtk_nor_rmw(sp, MTK_NOR_REG_BUSCFG, 0, MTK_NOR_4B_ADDR);
}
}
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
static bool need_bounce(struct mtk_nor *sp, const struct spi_mem_op *op)
{
return ((uintptr_t)op->data.buf.in & MTK_NOR_DMA_ALIGN_MASK);
}
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
static bool mtk_nor_match_read(const struct spi_mem_op *op)
{
int dummy = 0;
spi: Fixed division by zero warning The reason for dividing by zero is because the dummy bus width is zero, but if the dummy n bytes is zero, it indicates that there is no data transfer, so there is no need for calculation. Fixes: 7512eaf54190 ("spi: cadence-quadspi: Fix dummy cycle calculation when buswidth > 1") Signed-off-by: Yoshitaka Ikeda <ikeda@nskint.co.jp> Acked-by: Pratyush Yadav <p.yadav@ti.com> Link: https://lore.kernel.org/r/OSZPR01MB70049C8F56ED8902852DF97B8BD49@OSZPR01MB7004.jpnprd01.prod.outlook.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-09-08 08:29:12 +03:00
if (op->dummy.nbytes)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
dummy = op->dummy.nbytes * BITS_PER_BYTE / op->dummy.buswidth;
if ((op->data.buswidth == 2) || (op->data.buswidth == 4)) {
if (op->addr.buswidth == 1)
return dummy == 8;
else if (op->addr.buswidth == 2)
return dummy == 4;
else if (op->addr.buswidth == 4)
return dummy == 6;
} else if ((op->addr.buswidth == 1) && (op->data.buswidth == 1)) {
if (op->cmd.opcode == 0x03)
return dummy == 0;
else if (op->cmd.opcode == 0x0b)
return dummy == 8;
}
return false;
}
spi: spi-mtk-nor: add helper for checking prg mode ops op checking/resizing logic for the newly added mtk_nor_spi_mem_prg is more complicated. Add two helper functions for them: mtk_nor_match_prg: check whether an op is supported by prg mode. mtk_nor_adj_prg_size: adjust data size for mtk_nor_spi_mem_prg. mtk_nor_match_prg isn't called yet because supports_op is currently broken. It'll be used in the next fix commit. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:29 +03:00
static bool mtk_nor_match_prg(const struct spi_mem_op *op)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
{
spi: spi-mtk-nor: add helper for checking prg mode ops op checking/resizing logic for the newly added mtk_nor_spi_mem_prg is more complicated. Add two helper functions for them: mtk_nor_match_prg: check whether an op is supported by prg mode. mtk_nor_adj_prg_size: adjust data size for mtk_nor_spi_mem_prg. mtk_nor_match_prg isn't called yet because supports_op is currently broken. It'll be used in the next fix commit. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:29 +03:00
int tx_len, rx_len, prg_len, prg_left;
// prg mode is spi-only.
if ((op->cmd.buswidth > 1) || (op->addr.buswidth > 1) ||
(op->dummy.buswidth > 1) || (op->data.buswidth > 1))
return false;
tx_len = op->cmd.nbytes + op->addr.nbytes;
if (op->data.dir == SPI_MEM_DATA_OUT) {
// count dummy bytes only if we need to write data after it
tx_len += op->dummy.nbytes;
// leave at least one byte for data
if (tx_len > MTK_NOR_REG_PRGDATA_MAX)
return false;
// if there's no addr, meaning adjust_op_size is impossible,
// check data length as well.
if ((!op->addr.nbytes) &&
(tx_len + op->data.nbytes > MTK_NOR_REG_PRGDATA_MAX + 1))
return false;
} else if (op->data.dir == SPI_MEM_DATA_IN) {
if (tx_len > MTK_NOR_REG_PRGDATA_MAX + 1)
return false;
rx_len = op->data.nbytes;
prg_left = MTK_NOR_PRG_CNT_MAX / 8 - tx_len - op->dummy.nbytes;
if (prg_left > MTK_NOR_REG_SHIFT_MAX + 1)
prg_left = MTK_NOR_REG_SHIFT_MAX + 1;
if (rx_len > prg_left) {
if (!op->addr.nbytes)
return false;
rx_len = prg_left;
}
prg_len = tx_len + op->dummy.nbytes + rx_len;
if (prg_len > MTK_NOR_PRG_CNT_MAX / 8)
return false;
} else {
prg_len = tx_len + op->dummy.nbytes;
if (prg_len > MTK_NOR_PRG_CNT_MAX / 8)
return false;
}
return true;
}
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
spi: spi-mtk-nor: add helper for checking prg mode ops op checking/resizing logic for the newly added mtk_nor_spi_mem_prg is more complicated. Add two helper functions for them: mtk_nor_match_prg: check whether an op is supported by prg mode. mtk_nor_adj_prg_size: adjust data size for mtk_nor_spi_mem_prg. mtk_nor_match_prg isn't called yet because supports_op is currently broken. It'll be used in the next fix commit. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:29 +03:00
static void mtk_nor_adj_prg_size(struct spi_mem_op *op)
{
int tx_len, tx_left, prg_left;
tx_len = op->cmd.nbytes + op->addr.nbytes;
if (op->data.dir == SPI_MEM_DATA_OUT) {
tx_len += op->dummy.nbytes;
tx_left = MTK_NOR_REG_PRGDATA_MAX + 1 - tx_len;
if (op->data.nbytes > tx_left)
op->data.nbytes = tx_left;
} else if (op->data.dir == SPI_MEM_DATA_IN) {
prg_left = MTK_NOR_PRG_CNT_MAX / 8 - tx_len - op->dummy.nbytes;
if (prg_left > MTK_NOR_REG_SHIFT_MAX + 1)
prg_left = MTK_NOR_REG_SHIFT_MAX + 1;
if (op->data.nbytes > prg_left)
op->data.nbytes = prg_left;
}
}
static int mtk_nor_adjust_op_size(struct spi_mem *mem, struct spi_mem_op *op)
{
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
struct mtk_nor *sp = spi_controller_get_devdata(mem->spi->master);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if (!op->data.nbytes)
return 0;
if ((op->addr.nbytes == 3) || (op->addr.nbytes == 4)) {
if ((op->data.dir == SPI_MEM_DATA_IN) &&
mtk_nor_match_read(op)) {
spi: spi-mtk-nor: fix timeout calculation overflow CLK_TO_US macro is used to calculate potential transfer time for various timeout handling. However it overflows on transfer bigger than 512 bytes because it first did (len * 8 * 1000000). This controller typically operates at 45MHz. This patch did 2 things: 1. calculate clock / 1000000 first 2. add a 4M transfer size cap so that the final timeout in DMA reading doesn't overflow Fixes: 881d1ee9fe81f ("spi: add support for mediatek spi-nor controller") Cc: <stable@vger.kernel.org> Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200922114905.2942859-1-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-22 14:49:02 +03:00
// limit size to prevent timeout calculation overflow
if (op->data.nbytes > 0x400000)
op->data.nbytes = 0x400000;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if ((op->addr.val & MTK_NOR_DMA_ALIGN_MASK) ||
(op->data.nbytes < MTK_NOR_DMA_ALIGN))
op->data.nbytes = 1;
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
else if (!need_bounce(sp, op))
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
op->data.nbytes &= ~MTK_NOR_DMA_ALIGN_MASK;
else if (op->data.nbytes > MTK_NOR_BOUNCE_BUF_SIZE)
op->data.nbytes = MTK_NOR_BOUNCE_BUF_SIZE;
return 0;
} else if (op->data.dir == SPI_MEM_DATA_OUT) {
if (op->data.nbytes >= MTK_NOR_PP_SIZE)
op->data.nbytes = MTK_NOR_PP_SIZE;
else
op->data.nbytes = 1;
return 0;
}
}
spi: spi-mtk-nor: add helper for checking prg mode ops op checking/resizing logic for the newly added mtk_nor_spi_mem_prg is more complicated. Add two helper functions for them: mtk_nor_match_prg: check whether an op is supported by prg mode. mtk_nor_adj_prg_size: adjust data size for mtk_nor_spi_mem_prg. mtk_nor_match_prg isn't called yet because supports_op is currently broken. It'll be used in the next fix commit. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:29 +03:00
mtk_nor_adj_prg_size(op);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
return 0;
}
static bool mtk_nor_supports_op(struct spi_mem *mem,
const struct spi_mem_op *op)
{
spi: spi-mtk-nor: fix op checks in supports_op commit a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") tries to inverse the logic of supports_op when adding spi_mem_default_supports_op check, but it didn't get it done properly. There are two regressions introduced by this commit: 1. reading ops supported by program mode is rejected. 2. all ops with special controller routines are incorrectly further checked against program mode. This commits inverses the logic back: 1. check spi_mem_default_supports_op and reject unsupported ops first. 2. return true for ops with special controller routines. 3. check the left ops against controller program mode. Fixes: a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-4-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:30 +03:00
if (!spi_mem_default_supports_op(mem, op))
return false;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if (op->cmd.buswidth != 1)
return false;
if ((op->addr.nbytes == 3) || (op->addr.nbytes == 4)) {
spi: spi-mtk-nor: Fix checkpatch spacing error Fix checkpatch error: ERROR: space required before the open parenthesis '(' #295: FILE: spi-mtk-nor.c:295: + switch(op->data.dir) { Signed-off-by: Jay Fang <f.fangjian@huawei.com> Link: https://lore.kernel.org/r/1616566602-13894-6-git-send-email-f.fangjian@huawei.com Signed-off-by: Mark Brown <broonie@kernel.org>
2021-03-24 09:16:36 +03:00
switch (op->data.dir) {
spi: spi-mtk-nor: support standard spi properties Use default supports_op() to support spi-[rt]x-bus-width properties. And check dummy op's byte length instead of its bus width for output. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20200826091852.519138-1-ikjn@chromium.org Signed-off-by: Mark Brown <broonie@kernel.org>
2020-08-26 12:18:52 +03:00
case SPI_MEM_DATA_IN:
spi: spi-mtk-nor: fix op checks in supports_op commit a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") tries to inverse the logic of supports_op when adding spi_mem_default_supports_op check, but it didn't get it done properly. There are two regressions introduced by this commit: 1. reading ops supported by program mode is rejected. 2. all ops with special controller routines are incorrectly further checked against program mode. This commits inverses the logic back: 1. check spi_mem_default_supports_op and reject unsupported ops first. 2. return true for ops with special controller routines. 3. check the left ops against controller program mode. Fixes: a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-4-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:30 +03:00
if (mtk_nor_match_read(op))
return true;
spi: spi-mtk-nor: support standard spi properties Use default supports_op() to support spi-[rt]x-bus-width properties. And check dummy op's byte length instead of its bus width for output. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20200826091852.519138-1-ikjn@chromium.org Signed-off-by: Mark Brown <broonie@kernel.org>
2020-08-26 12:18:52 +03:00
break;
case SPI_MEM_DATA_OUT:
spi: spi-mtk-nor: fix op checks in supports_op commit a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") tries to inverse the logic of supports_op when adding spi_mem_default_supports_op check, but it didn't get it done properly. There are two regressions introduced by this commit: 1. reading ops supported by program mode is rejected. 2. all ops with special controller routines are incorrectly further checked against program mode. This commits inverses the logic back: 1. check spi_mem_default_supports_op and reject unsupported ops first. 2. return true for ops with special controller routines. 3. check the left ops against controller program mode. Fixes: a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-4-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:30 +03:00
if ((op->addr.buswidth == 1) &&
(op->dummy.nbytes == 0) &&
(op->data.buswidth == 1))
return true;
spi: spi-mtk-nor: support standard spi properties Use default supports_op() to support spi-[rt]x-bus-width properties. And check dummy op's byte length instead of its bus width for output. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20200826091852.519138-1-ikjn@chromium.org Signed-off-by: Mark Brown <broonie@kernel.org>
2020-08-26 12:18:52 +03:00
break;
default:
break;
}
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
spi: spi-mtk-nor: support standard spi properties Use default supports_op() to support spi-[rt]x-bus-width properties. And check dummy op's byte length instead of its bus width for output. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20200826091852.519138-1-ikjn@chromium.org Signed-off-by: Mark Brown <broonie@kernel.org>
2020-08-26 12:18:52 +03:00
spi: spi-mtk-nor: fix op checks in supports_op commit a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") tries to inverse the logic of supports_op when adding spi_mem_default_supports_op check, but it didn't get it done properly. There are two regressions introduced by this commit: 1. reading ops supported by program mode is rejected. 2. all ops with special controller routines are incorrectly further checked against program mode. This commits inverses the logic back: 1. check spi_mem_default_supports_op and reject unsupported ops first. 2. return true for ops with special controller routines. 3. check the left ops against controller program mode. Fixes: a59b2c7c56bf7 ("spi: spi-mtk-nor: support standard spi properties") Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-4-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:30 +03:00
return mtk_nor_match_prg(op);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
static void mtk_nor_setup_bus(struct mtk_nor *sp, const struct spi_mem_op *op)
{
u32 reg = 0;
if (op->addr.nbytes == 4)
reg |= MTK_NOR_4B_ADDR;
if (op->data.buswidth == 4) {
reg |= MTK_NOR_QUAD_READ;
writeb(op->cmd.opcode, sp->base + MTK_NOR_REG_PRGDATA(4));
if (op->addr.buswidth == 4)
reg |= MTK_NOR_QUAD_ADDR;
} else if (op->data.buswidth == 2) {
reg |= MTK_NOR_DUAL_READ;
writeb(op->cmd.opcode, sp->base + MTK_NOR_REG_PRGDATA(3));
if (op->addr.buswidth == 2)
reg |= MTK_NOR_DUAL_ADDR;
} else {
if (op->cmd.opcode == 0x0b)
mtk_nor_rmw(sp, MTK_NOR_REG_CFG1, MTK_NOR_FAST_READ, 0);
else
mtk_nor_rmw(sp, MTK_NOR_REG_CFG1, 0, MTK_NOR_FAST_READ);
}
mtk_nor_rmw(sp, MTK_NOR_REG_BUSCFG, reg, MTK_NOR_BUS_MODE_MASK);
}
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
static int mtk_nor_dma_exec(struct mtk_nor *sp, u32 from, unsigned int length,
dma_addr_t dma_addr)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
{
int ret = 0;
ulong delay;
u32 reg;
writel(from, sp->base + MTK_NOR_REG_DMA_FADR);
writel(dma_addr, sp->base + MTK_NOR_REG_DMA_DADR);
writel(dma_addr + length, sp->base + MTK_NOR_REG_DMA_END_DADR);
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
if (sp->high_dma) {
writel(upper_32_bits(dma_addr),
sp->base + MTK_NOR_REG_DMA_DADR_HB);
writel(upper_32_bits(dma_addr + length),
sp->base + MTK_NOR_REG_DMA_END_DADR_HB);
}
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if (sp->has_irq) {
reinit_completion(&sp->op_done);
mtk_nor_rmw(sp, MTK_NOR_REG_IRQ_EN, MTK_NOR_IRQ_DMA, 0);
}
mtk_nor_rmw(sp, MTK_NOR_REG_DMA_CTL, MTK_NOR_DMA_START, 0);
delay = CLK_TO_US(sp, (length + 5) * BITS_PER_BYTE);
if (sp->has_irq) {
if (!wait_for_completion_timeout(&sp->op_done,
(delay + 1) * 100))
ret = -ETIMEDOUT;
} else {
ret = readl_poll_timeout(sp->base + MTK_NOR_REG_DMA_CTL, reg,
!(reg & MTK_NOR_DMA_START), delay / 3,
(delay + 1) * 100);
}
if (ret < 0)
dev_err(sp->dev, "dma read timeout.\n");
return ret;
}
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
static int mtk_nor_read_bounce(struct mtk_nor *sp, const struct spi_mem_op *op)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
{
unsigned int rdlen;
int ret;
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
if (op->data.nbytes & MTK_NOR_DMA_ALIGN_MASK)
rdlen = (op->data.nbytes + MTK_NOR_DMA_ALIGN) & ~MTK_NOR_DMA_ALIGN_MASK;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
else
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
rdlen = op->data.nbytes;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
ret = mtk_nor_dma_exec(sp, op->addr.val, rdlen, sp->buffer_dma);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
if (!ret)
memcpy(op->data.buf.in, sp->buffer, op->data.nbytes);
return ret;
}
static int mtk_nor_read_dma(struct mtk_nor *sp, const struct spi_mem_op *op)
{
int ret;
dma_addr_t dma_addr;
if (need_bounce(sp, op))
return mtk_nor_read_bounce(sp, op);
dma_addr = dma_map_single(sp->dev, op->data.buf.in,
op->data.nbytes, DMA_FROM_DEVICE);
if (dma_mapping_error(sp->dev, dma_addr))
return -EINVAL;
ret = mtk_nor_dma_exec(sp, op->addr.val, op->data.nbytes, dma_addr);
dma_unmap_single(sp->dev, dma_addr, op->data.nbytes, DMA_FROM_DEVICE);
return ret;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
static int mtk_nor_read_pio(struct mtk_nor *sp, const struct spi_mem_op *op)
{
u8 *buf = op->data.buf.in;
int ret;
ret = mtk_nor_cmd_exec(sp, MTK_NOR_CMD_READ, 6 * BITS_PER_BYTE);
if (!ret)
buf[0] = readb(sp->base + MTK_NOR_REG_RDATA);
return ret;
}
static int mtk_nor_write_buffer_enable(struct mtk_nor *sp)
{
int ret;
u32 val;
if (sp->wbuf_en)
return 0;
val = readl(sp->base + MTK_NOR_REG_CFG2);
writel(val | MTK_NOR_WR_BUF_EN, sp->base + MTK_NOR_REG_CFG2);
ret = readl_poll_timeout(sp->base + MTK_NOR_REG_CFG2, val,
val & MTK_NOR_WR_BUF_EN, 0, 10000);
if (!ret)
sp->wbuf_en = true;
return ret;
}
static int mtk_nor_write_buffer_disable(struct mtk_nor *sp)
{
int ret;
u32 val;
if (!sp->wbuf_en)
return 0;
val = readl(sp->base + MTK_NOR_REG_CFG2);
writel(val & ~MTK_NOR_WR_BUF_EN, sp->base + MTK_NOR_REG_CFG2);
ret = readl_poll_timeout(sp->base + MTK_NOR_REG_CFG2, val,
!(val & MTK_NOR_WR_BUF_EN), 0, 10000);
if (!ret)
sp->wbuf_en = false;
return ret;
}
static int mtk_nor_pp_buffered(struct mtk_nor *sp, const struct spi_mem_op *op)
{
const u8 *buf = op->data.buf.out;
u32 val;
int ret, i;
ret = mtk_nor_write_buffer_enable(sp);
if (ret < 0)
return ret;
for (i = 0; i < op->data.nbytes; i += 4) {
val = buf[i + 3] << 24 | buf[i + 2] << 16 | buf[i + 1] << 8 |
buf[i];
writel(val, sp->base + MTK_NOR_REG_PP_DATA);
}
return mtk_nor_cmd_exec(sp, MTK_NOR_CMD_WRITE,
(op->data.nbytes + 5) * BITS_PER_BYTE);
}
static int mtk_nor_pp_unbuffered(struct mtk_nor *sp,
const struct spi_mem_op *op)
{
const u8 *buf = op->data.buf.out;
int ret;
ret = mtk_nor_write_buffer_disable(sp);
if (ret < 0)
return ret;
writeb(buf[0], sp->base + MTK_NOR_REG_WDATA);
return mtk_nor_cmd_exec(sp, MTK_NOR_CMD_WRITE, 6 * BITS_PER_BYTE);
}
spi: spi-mtk-nor: make use of full capability of prg mode "program" mode on this controller can trigger up to 56 bits of data shifting. During the operation, data in PRGDATA[0-5] will be shifted out from MOSI, and data from MISO will be continuously filling SHREG[0-9]. Currently this mode is used to implement transfer_one_message for 6-byte full-duplex transfer, but it can execute a transfer for up-to 7 bytes as long as the last byte is read only. transfer_one_message is expected to perform full-duplex transfer, instead of transfer with specific format. mtk_nor_spi_mem_prg is added here to use this extra byte. Newer version of this controller can trigger longer data shifting with shift bytes more than PRGDATA_MAX + SHREG_MAX. This patch is implemented with that in mind and it checks against both SHREG_MAX and PRG_CNT_MAX for future support of new controllers. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-2-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:28 +03:00
static int mtk_nor_spi_mem_prg(struct mtk_nor *sp, const struct spi_mem_op *op)
{
int rx_len = 0;
int reg_offset = MTK_NOR_REG_PRGDATA_MAX;
int tx_len, prg_len;
int i, ret;
void __iomem *reg;
u8 bufbyte;
tx_len = op->cmd.nbytes + op->addr.nbytes;
// count dummy bytes only if we need to write data after it
if (op->data.dir == SPI_MEM_DATA_OUT)
tx_len += op->dummy.nbytes + op->data.nbytes;
else if (op->data.dir == SPI_MEM_DATA_IN)
rx_len = op->data.nbytes;
prg_len = op->cmd.nbytes + op->addr.nbytes + op->dummy.nbytes +
op->data.nbytes;
// an invalid op may reach here if the caller calls exec_op without
// adjust_op_size. return -EINVAL instead of -ENOTSUPP so that
// spi-mem won't try this op again with generic spi transfers.
if ((tx_len > MTK_NOR_REG_PRGDATA_MAX + 1) ||
(rx_len > MTK_NOR_REG_SHIFT_MAX + 1) ||
(prg_len > MTK_NOR_PRG_CNT_MAX / 8))
return -EINVAL;
// fill tx data
for (i = op->cmd.nbytes; i > 0; i--, reg_offset--) {
reg = sp->base + MTK_NOR_REG_PRGDATA(reg_offset);
bufbyte = (op->cmd.opcode >> ((i - 1) * BITS_PER_BYTE)) & 0xff;
writeb(bufbyte, reg);
}
for (i = op->addr.nbytes; i > 0; i--, reg_offset--) {
reg = sp->base + MTK_NOR_REG_PRGDATA(reg_offset);
bufbyte = (op->addr.val >> ((i - 1) * BITS_PER_BYTE)) & 0xff;
writeb(bufbyte, reg);
}
if (op->data.dir == SPI_MEM_DATA_OUT) {
for (i = 0; i < op->dummy.nbytes; i++, reg_offset--) {
reg = sp->base + MTK_NOR_REG_PRGDATA(reg_offset);
writeb(0, reg);
}
for (i = 0; i < op->data.nbytes; i++, reg_offset--) {
reg = sp->base + MTK_NOR_REG_PRGDATA(reg_offset);
writeb(((const u8 *)(op->data.buf.out))[i], reg);
}
}
for (; reg_offset >= 0; reg_offset--) {
reg = sp->base + MTK_NOR_REG_PRGDATA(reg_offset);
writeb(0, reg);
}
// trigger op
writel(prg_len * BITS_PER_BYTE, sp->base + MTK_NOR_REG_PRG_CNT);
ret = mtk_nor_cmd_exec(sp, MTK_NOR_CMD_PROGRAM,
prg_len * BITS_PER_BYTE);
if (ret)
return ret;
// fetch read data
reg_offset = 0;
if (op->data.dir == SPI_MEM_DATA_IN) {
for (i = op->data.nbytes - 1; i >= 0; i--, reg_offset++) {
reg = sp->base + MTK_NOR_REG_SHIFT(reg_offset);
((u8 *)(op->data.buf.in))[i] = readb(reg);
}
}
return 0;
}
spi: spi-mtk-nor: make mtk_nor_exec_op() statuc Fix the following sparse warning: drivers/spi/spi-mtk-nor.c:394:5: warning: symbol 'mtk_nor_exec_op' was not declared. Should it be static? Reported-by: Hulk Robot <hulkci@huawei.com> Signed-off-by: Jason Yan <yanaijie@huawei.com> Link: https://lore.kernel.org/r/20200409085009.44971-1-yanaijie@huawei.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-04-09 11:50:09 +03:00
static int mtk_nor_exec_op(struct spi_mem *mem, const struct spi_mem_op *op)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
{
struct mtk_nor *sp = spi_controller_get_devdata(mem->spi->master);
int ret;
if ((op->data.nbytes == 0) ||
((op->addr.nbytes != 3) && (op->addr.nbytes != 4)))
spi: spi-mtk-nor: make use of full capability of prg mode "program" mode on this controller can trigger up to 56 bits of data shifting. During the operation, data in PRGDATA[0-5] will be shifted out from MOSI, and data from MISO will be continuously filling SHREG[0-9]. Currently this mode is used to implement transfer_one_message for 6-byte full-duplex transfer, but it can execute a transfer for up-to 7 bytes as long as the last byte is read only. transfer_one_message is expected to perform full-duplex transfer, instead of transfer with specific format. mtk_nor_spi_mem_prg is added here to use this extra byte. Newer version of this controller can trigger longer data shifting with shift bytes more than PRGDATA_MAX + SHREG_MAX. This patch is implemented with that in mind and it checks against both SHREG_MAX and PRG_CNT_MAX for future support of new controllers. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-2-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:28 +03:00
return mtk_nor_spi_mem_prg(sp, op);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if (op->data.dir == SPI_MEM_DATA_OUT) {
mtk_nor_set_addr(sp, op);
writeb(op->cmd.opcode, sp->base + MTK_NOR_REG_PRGDATA0);
if (op->data.nbytes == MTK_NOR_PP_SIZE)
return mtk_nor_pp_buffered(sp, op);
return mtk_nor_pp_unbuffered(sp, op);
}
if ((op->data.dir == SPI_MEM_DATA_IN) && mtk_nor_match_read(op)) {
ret = mtk_nor_write_buffer_disable(sp);
if (ret < 0)
return ret;
mtk_nor_setup_bus(sp, op);
if (op->data.nbytes == 1) {
mtk_nor_set_addr(sp, op);
return mtk_nor_read_pio(sp, op);
} else {
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
return mtk_nor_read_dma(sp, op);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
}
spi: spi-mtk-nor: make use of full capability of prg mode "program" mode on this controller can trigger up to 56 bits of data shifting. During the operation, data in PRGDATA[0-5] will be shifted out from MOSI, and data from MISO will be continuously filling SHREG[0-9]. Currently this mode is used to implement transfer_one_message for 6-byte full-duplex transfer, but it can execute a transfer for up-to 7 bytes as long as the last byte is read only. transfer_one_message is expected to perform full-duplex transfer, instead of transfer with specific format. mtk_nor_spi_mem_prg is added here to use this extra byte. Newer version of this controller can trigger longer data shifting with shift bytes more than PRGDATA_MAX + SHREG_MAX. This patch is implemented with that in mind and it checks against both SHREG_MAX and PRG_CNT_MAX for future support of new controllers. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200924152730.733243-2-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-09-24 18:27:28 +03:00
return mtk_nor_spi_mem_prg(sp, op);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
static int mtk_nor_setup(struct spi_device *spi)
{
struct mtk_nor *sp = spi_controller_get_devdata(spi->master);
if (spi->max_speed_hz && (spi->max_speed_hz < sp->spi_freq)) {
dev_err(&spi->dev, "spi clock should be %u Hz.\n",
sp->spi_freq);
return -EINVAL;
}
spi->max_speed_hz = sp->spi_freq;
return 0;
}
static int mtk_nor_transfer_one_message(struct spi_controller *master,
struct spi_message *m)
{
struct mtk_nor *sp = spi_controller_get_devdata(master);
struct spi_transfer *t = NULL;
unsigned long trx_len = 0;
int stat = 0;
int reg_offset = MTK_NOR_REG_PRGDATA_MAX;
void __iomem *reg;
const u8 *txbuf;
u8 *rxbuf;
int i;
list_for_each_entry(t, &m->transfers, transfer_list) {
txbuf = t->tx_buf;
for (i = 0; i < t->len; i++, reg_offset--) {
reg = sp->base + MTK_NOR_REG_PRGDATA(reg_offset);
if (txbuf)
writeb(txbuf[i], reg);
else
writeb(0, reg);
}
trx_len += t->len;
}
writel(trx_len * BITS_PER_BYTE, sp->base + MTK_NOR_REG_PRG_CNT);
stat = mtk_nor_cmd_exec(sp, MTK_NOR_CMD_PROGRAM,
trx_len * BITS_PER_BYTE);
if (stat < 0)
goto msg_done;
reg_offset = trx_len - 1;
list_for_each_entry(t, &m->transfers, transfer_list) {
rxbuf = t->rx_buf;
for (i = 0; i < t->len; i++, reg_offset--) {
reg = sp->base + MTK_NOR_REG_SHIFT(reg_offset);
if (rxbuf)
rxbuf[i] = readb(reg);
}
}
m->actual_length = trx_len;
msg_done:
m->status = stat;
spi_finalize_current_message(master);
return 0;
}
static void mtk_nor_disable_clk(struct mtk_nor *sp)
{
clk_disable_unprepare(sp->spi_clk);
clk_disable_unprepare(sp->ctlr_clk);
spi: spi-mtk-nor: add axi clock control for MT8192 spi-nor MT8192 spi-nor is an independent sub system, we need extra control axi bus clock for it. Add support for the additional axi clock to allow it to be configured appropriately. Signed-off-by: bayi cheng <bayi.cheng@mediatek.com> Tested-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/1605084902-13151-2-git-send-email-bayi.cheng@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-11-11 11:55:02 +03:00
clk_disable_unprepare(sp->axi_clk);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
static int mtk_nor_enable_clk(struct mtk_nor *sp)
{
int ret;
ret = clk_prepare_enable(sp->spi_clk);
if (ret)
return ret;
ret = clk_prepare_enable(sp->ctlr_clk);
if (ret) {
clk_disable_unprepare(sp->spi_clk);
return ret;
}
spi: spi-mtk-nor: add axi clock control for MT8192 spi-nor MT8192 spi-nor is an independent sub system, we need extra control axi bus clock for it. Add support for the additional axi clock to allow it to be configured appropriately. Signed-off-by: bayi cheng <bayi.cheng@mediatek.com> Tested-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/1605084902-13151-2-git-send-email-bayi.cheng@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-11-11 11:55:02 +03:00
ret = clk_prepare_enable(sp->axi_clk);
if (ret) {
clk_disable_unprepare(sp->spi_clk);
clk_disable_unprepare(sp->ctlr_clk);
return ret;
}
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
return 0;
}
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
static void mtk_nor_init(struct mtk_nor *sp)
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
{
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
writel(0, sp->base + MTK_NOR_REG_IRQ_EN);
writel(MTK_NOR_IRQ_MASK, sp->base + MTK_NOR_REG_IRQ_STAT);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
writel(MTK_NOR_ENABLE_SF_CMD, sp->base + MTK_NOR_REG_WP);
mtk_nor_rmw(sp, MTK_NOR_REG_CFG2, MTK_NOR_WR_CUSTOM_OP_EN, 0);
mtk_nor_rmw(sp, MTK_NOR_REG_CFG3,
MTK_NOR_DISABLE_WREN | MTK_NOR_DISABLE_SR_POLL, 0);
}
static irqreturn_t mtk_nor_irq_handler(int irq, void *data)
{
struct mtk_nor *sp = data;
u32 irq_status, irq_enabled;
irq_status = readl(sp->base + MTK_NOR_REG_IRQ_STAT);
irq_enabled = readl(sp->base + MTK_NOR_REG_IRQ_EN);
// write status back to clear interrupt
writel(irq_status, sp->base + MTK_NOR_REG_IRQ_STAT);
if (!(irq_status & irq_enabled))
return IRQ_NONE;
if (irq_status & MTK_NOR_IRQ_DMA) {
complete(&sp->op_done);
writel(0, sp->base + MTK_NOR_REG_IRQ_EN);
}
return IRQ_HANDLED;
}
static size_t mtk_max_msg_size(struct spi_device *spi)
{
return MTK_NOR_PRG_MAX_SIZE;
}
static const struct spi_controller_mem_ops mtk_nor_mem_ops = {
.adjust_op_size = mtk_nor_adjust_op_size,
.supports_op = mtk_nor_supports_op,
.exec_op = mtk_nor_exec_op
};
static const struct of_device_id mtk_nor_match[] = {
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
{ .compatible = "mediatek,mt8192-nor", .data = (void *)36 },
{ .compatible = "mediatek,mt8173-nor", .data = (void *)32 },
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
{ /* sentinel */ }
};
MODULE_DEVICE_TABLE(of, mtk_nor_match);
static int mtk_nor_probe(struct platform_device *pdev)
{
struct spi_controller *ctlr;
struct mtk_nor *sp;
void __iomem *base;
spi: spi-mtk-nor: add axi clock control for MT8192 spi-nor MT8192 spi-nor is an independent sub system, we need extra control axi bus clock for it. Add support for the additional axi clock to allow it to be configured appropriately. Signed-off-by: bayi cheng <bayi.cheng@mediatek.com> Tested-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/1605084902-13151-2-git-send-email-bayi.cheng@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-11-11 11:55:02 +03:00
struct clk *spi_clk, *ctlr_clk, *axi_clk;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
int ret, irq;
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
unsigned long dma_bits;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
base = devm_platform_ioremap_resource(pdev, 0);
if (IS_ERR(base))
return PTR_ERR(base);
spi_clk = devm_clk_get(&pdev->dev, "spi");
if (IS_ERR(spi_clk))
return PTR_ERR(spi_clk);
ctlr_clk = devm_clk_get(&pdev->dev, "sf");
if (IS_ERR(ctlr_clk))
return PTR_ERR(ctlr_clk);
spi: spi-mtk-nor: add axi clock control for MT8192 spi-nor MT8192 spi-nor is an independent sub system, we need extra control axi bus clock for it. Add support for the additional axi clock to allow it to be configured appropriately. Signed-off-by: bayi cheng <bayi.cheng@mediatek.com> Tested-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/1605084902-13151-2-git-send-email-bayi.cheng@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-11-11 11:55:02 +03:00
axi_clk = devm_clk_get_optional(&pdev->dev, "axi");
if (IS_ERR(axi_clk))
return PTR_ERR(axi_clk);
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
dma_bits = (unsigned long)of_device_get_match_data(&pdev->dev);
if (dma_set_mask_and_coherent(&pdev->dev, DMA_BIT_MASK(dma_bits))) {
dev_err(&pdev->dev, "failed to set dma mask(%lu)\n", dma_bits);
return -EINVAL;
}
spi: spi-mtk-nor: Don't leak SPI master in probe error path If the call to devm_spi_register_controller() fails on probe of the MediaTek SPI NOR driver, the spi_controller struct is erroneously not freed. Since commit a1daaa991ed1 ("spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer"), the same happens if the call to dmam_alloc_coherent() fails. Since commit 3bfd9103c7af ("spi: spi-mtk-nor: Add power management support"), the same happens if the call to mtk_nor_enable_clk() fails. Fix by switching over to the new devm_spi_alloc_master() helper. Fixes: 881d1ee9fe81 ("spi: add support for mediatek spi-nor controller") Signed-off-by: Lukas Wunner <lukas@wunner.de> Reviewed-by: Ikjoon Jang <ikjn@chromium.org> Cc: <stable@vger.kernel.org> # v5.7+: 5e844cc37a5c: spi: Introduce device-managed SPI controller allocation Cc: <stable@vger.kernel.org> # v5.7+ Cc: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/d5b9f0289465394e73dedb8ec51e180a8f1dffc9.1607286887.git.lukas@wunner.de Signed-off-by: Mark Brown <broonie@kernel.org>
2020-12-07 11:17:08 +03:00
ctlr = devm_spi_alloc_master(&pdev->dev, sizeof(*sp));
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if (!ctlr) {
dev_err(&pdev->dev, "failed to allocate spi controller\n");
return -ENOMEM;
}
ctlr->bits_per_word_mask = SPI_BPW_MASK(8);
ctlr->dev.of_node = pdev->dev.of_node;
ctlr->max_message_size = mtk_max_msg_size;
ctlr->mem_ops = &mtk_nor_mem_ops;
ctlr->mode_bits = SPI_RX_DUAL | SPI_RX_QUAD | SPI_TX_DUAL | SPI_TX_QUAD;
ctlr->num_chipselect = 1;
ctlr->setup = mtk_nor_setup;
ctlr->transfer_one_message = mtk_nor_transfer_one_message;
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
ctlr->auto_runtime_pm = true;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
dev_set_drvdata(&pdev->dev, ctlr);
sp = spi_controller_get_devdata(ctlr);
sp->base = base;
sp->has_irq = false;
sp->wbuf_en = false;
sp->ctlr = ctlr;
sp->dev = &pdev->dev;
sp->spi_clk = spi_clk;
sp->ctlr_clk = ctlr_clk;
spi: spi-mtk-nor: add axi clock control for MT8192 spi-nor MT8192 spi-nor is an independent sub system, we need extra control axi bus clock for it. Add support for the additional axi clock to allow it to be configured appropriately. Signed-off-by: bayi cheng <bayi.cheng@mediatek.com> Tested-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/1605084902-13151-2-git-send-email-bayi.cheng@mediatek.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-11-11 11:55:02 +03:00
sp->axi_clk = axi_clk;
spi: spi-mtk-nor: support 36bit dma addressing This patch enables 36bit dma address support to spi-mtk-nor. Currently this is enabled only for mt8192-nor. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.3.Id1cb208392928afc7ceed4de06924243c7858cd0@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:04 +03:00
sp->high_dma = (dma_bits > 32);
spi: spi-mtk-nor: use dma_alloc_coherent() for bounce buffer Use dma_alloc_coherent() for bounce buffer instead of kmalloc() to make sure the bounce buffer to be allocated within its DMAable range. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.2.I06cb65401ab5ad63ea30c4788d26633928d80f38@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:03 +03:00
sp->buffer = dmam_alloc_coherent(&pdev->dev,
MTK_NOR_BOUNCE_BUF_SIZE + MTK_NOR_DMA_ALIGN,
&sp->buffer_dma, GFP_KERNEL);
if (!sp->buffer)
return -ENOMEM;
if ((uintptr_t)sp->buffer & MTK_NOR_DMA_ALIGN_MASK) {
dev_err(sp->dev, "misaligned allocation of internal buffer.\n");
return -ENOMEM;
}
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
ret = mtk_nor_enable_clk(sp);
if (ret < 0)
return ret;
sp->spi_freq = clk_get_rate(sp->spi_clk);
mtk_nor_init(sp);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
irq = platform_get_irq_optional(pdev, 0);
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
if (irq < 0) {
dev_warn(sp->dev, "IRQ not available.");
} else {
ret = devm_request_irq(sp->dev, irq, mtk_nor_irq_handler, 0,
pdev->name, sp);
if (ret < 0) {
dev_warn(sp->dev, "failed to request IRQ.");
} else {
init_completion(&sp->op_done);
sp->has_irq = true;
}
}
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
pm_runtime_set_autosuspend_delay(&pdev->dev, -1);
pm_runtime_use_autosuspend(&pdev->dev);
pm_runtime_set_active(&pdev->dev);
pm_runtime_enable(&pdev->dev);
pm_runtime_get_noresume(&pdev->dev);
ret = devm_spi_register_controller(&pdev->dev, ctlr);
if (ret < 0)
goto err_probe;
pm_runtime_mark_last_busy(&pdev->dev);
pm_runtime_put_autosuspend(&pdev->dev);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
dev_info(&pdev->dev, "spi frequency: %d Hz\n", sp->spi_freq);
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
return 0;
err_probe:
pm_runtime_disable(&pdev->dev);
pm_runtime_set_suspended(&pdev->dev);
pm_runtime_dont_use_autosuspend(&pdev->dev);
mtk_nor_disable_clk(sp);
return ret;
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
}
static int mtk_nor_remove(struct platform_device *pdev)
{
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
struct spi_controller *ctlr = dev_get_drvdata(&pdev->dev);
struct mtk_nor *sp = spi_controller_get_devdata(ctlr);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
pm_runtime_disable(&pdev->dev);
pm_runtime_set_suspended(&pdev->dev);
pm_runtime_dont_use_autosuspend(&pdev->dev);
mtk_nor_disable_clk(sp);
return 0;
}
static int __maybe_unused mtk_nor_runtime_suspend(struct device *dev)
{
struct spi_controller *ctlr = dev_get_drvdata(dev);
struct mtk_nor *sp = spi_controller_get_devdata(ctlr);
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
mtk_nor_disable_clk(sp);
return 0;
}
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
static int __maybe_unused mtk_nor_runtime_resume(struct device *dev)
{
struct spi_controller *ctlr = dev_get_drvdata(dev);
struct mtk_nor *sp = spi_controller_get_devdata(ctlr);
return mtk_nor_enable_clk(sp);
}
static int __maybe_unused mtk_nor_suspend(struct device *dev)
{
return pm_runtime_force_suspend(dev);
}
static int __maybe_unused mtk_nor_resume(struct device *dev)
{
return pm_runtime_force_resume(dev);
}
static const struct dev_pm_ops mtk_nor_pm_ops = {
SET_RUNTIME_PM_OPS(mtk_nor_runtime_suspend,
mtk_nor_runtime_resume, NULL)
SET_SYSTEM_SLEEP_PM_OPS(mtk_nor_suspend, mtk_nor_resume)
};
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
static struct platform_driver mtk_nor_driver = {
.driver = {
.name = DRIVER_NAME,
.of_match_table = mtk_nor_match,
spi: spi-mtk-nor: Add power management support This patch adds dev_pm_ops to mtk-nor to support suspend/resume, auto suspend delay is set to -1 by default. Accessing registers are only permitted after its clock is enabled to deal with unknown state of operating clk at probe time. Signed-off-by: Ikjoon Jang <ikjn@chromium.org> Link: https://lore.kernel.org/r/20201006155010.v5.4.I68983b582d949a91866163bab588ff3c2a0d0275@changeid Signed-off-by: Mark Brown <broonie@kernel.org>
2020-10-06 10:54:05 +03:00
.pm = &mtk_nor_pm_ops,
spi: add support for mediatek spi-nor controller This is a driver for mtk spi-nor controller using spi-mem interface. The same controller already has limited support provided by mtk-quadspi driver under spi-nor framework and this new driver is a replacement for the old one. Comparing to the old driver, this driver has following advantages: 1. It can handle any full-duplex spi transfer up to 6 bytes, and this is implemented using generic spi interface. 2. It take account into command opcode properly. The reading routine in this controller can only use 0x03 or 0x0b as opcode on 1-1-1 transfers, but old driver doesn't implement this properly. This driver checks supported opcode explicitly and use (1) to perform unmatched operations. 3. It properly handles SFDP reading. Old driver can't read SFDP due to the bug mentioned in (2). 4. It can do 1-2-2 and 1-4-4 fast reading on spi-nor. These two ops requires parsing SFDP, which isn't possible in old driver. And the old driver is only flagged to support 1-1-2 mode. 5. It takes advantage of the DMA feature in this controller for long reads and supports IRQ on DMA requests to free cpu cycles from polling status registers on long DMA reading. It achieves up to 17.5MB/s reading speed (1-4-4 mode) which is way faster than the old one. IRQ is implemented as optional to maintain backward compatibility. Signed-off-by: Chuanhong Guo <gch981213@gmail.com> Link: https://lore.kernel.org/r/20200306085052.28258-3-gch981213@gmail.com Signed-off-by: Mark Brown <broonie@kernel.org>
2020-03-06 11:50:50 +03:00
},
.probe = mtk_nor_probe,
.remove = mtk_nor_remove,
};
module_platform_driver(mtk_nor_driver);
MODULE_DESCRIPTION("Mediatek SPI NOR controller driver");
MODULE_AUTHOR("Chuanhong Guo <gch981213@gmail.com>");
MODULE_LICENSE("GPL v2");
MODULE_ALIAS("platform:" DRIVER_NAME);
Reference in New Issue Copy Permalink