linux-mtd.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: Vignesh Raghavendra <vigneshr@ti.com>
To: "Ramuthevar,Vadivel MuruganX"
	<vadivel.muruganx.ramuthevar@linux.intel.com>,
	<linux-mtd@lists.infradead.org>
Cc: cheol.yong.kim@intel.com, devicetree@vger.kernel.org,
	tudor.ambarus@gmail.com, andriy.shevchenko@intel.com,
	boris.brezillon@free-electrons.com, richard@nod.at,
	qi-ming.wu@intel.com, linux-kernel@vger.kernel.org,
	david.oberhollenzer@sigma-star.at, miquel.raynal@bootlin.com,
	jwboyer@gmail.com, computersforpeace@gmail.com,
	dwmw2@infradead.org, cyrille.pitchen@atmel.com
Subject: Re: [PATCH v2 2/3] mtd: spi-nor: cadence-quadspi: disable DMA and DAC for Intel LGM
Date: Mon, 9 Sep 2019 11:35:26 +0530	[thread overview]
Message-ID: <2596ecd4-4ba6-ff7d-472f-1f6e664b4a97@ti.com> (raw)
In-Reply-To: <20190827035827.21024-3-vadivel.muruganx.ramuthevar@linux.intel.com>

Hi,

On 27/08/19 9:28 AM, Ramuthevar,Vadivel MuruganX wrote:
> From: Ramuthevar Vadivel Murugan <vadivel.muruganx.ramuthevar@linux.intel.com>
> 
> on Intel's Lightning Mountain(LGM) SoCs QSPI controller do not use
> Direct Memory Access(DMA) and Direct Access Controller(DAC).
> 
> This patch introduces to properly disable the DMA and DAC
> for data transfer instead it uses indirect data transfer.
> 
> Signed-off-by: Ramuthevar Vadivel Murugan <vadivel.muruganx.ramuthevar@linux.intel.com>
> ---
>  drivers/mtd/spi-nor/Kconfig           |  2 +-
>  drivers/mtd/spi-nor/cadence-quadspi.c | 21 ++++++++++++++++++---
>  2 files changed, 19 insertions(+), 4 deletions(-)
> 
> diff --git a/drivers/mtd/spi-nor/Kconfig b/drivers/mtd/spi-nor/Kconfig
> index 6de83277ce8b..ba2e372ae514 100644
> --- a/drivers/mtd/spi-nor/Kconfig
> +++ b/drivers/mtd/spi-nor/Kconfig
> @@ -34,7 +34,7 @@ config SPI_ASPEED_SMC
>  
>  config SPI_CADENCE_QUADSPI
>  	tristate "Cadence Quad SPI controller"
> -	depends on OF && (ARM || ARM64 || COMPILE_TEST)
> +	depends on OF && (ARM || ARM64 || COMPILE_TEST || X86)
>  	help
>  	  Enable support for the Cadence Quad SPI Flash controller.
>  
> diff --git a/drivers/mtd/spi-nor/cadence-quadspi.c b/drivers/mtd/spi-nor/cadence-quadspi.c
> index 67f15a1f16fd..69fa13e95110 100644
> --- a/drivers/mtd/spi-nor/cadence-quadspi.c
> +++ b/drivers/mtd/spi-nor/cadence-quadspi.c
> @@ -517,12 +517,16 @@ static int cqspi_indirect_read_execute(struct spi_nor *nor, u8 *rxbuf,
>  	struct cqspi_st *cqspi = f_pdata->cqspi;
>  	void __iomem *reg_base = cqspi->iobase;
>  	void __iomem *ahb_base = cqspi->ahb_base;
> +	u32 trigger_address = cqspi->trigger_address;
>  	unsigned int remaining = n_rx;
>  	unsigned int mod_bytes = n_rx % 4;
>  	unsigned int bytes_to_read = 0;
>  	u8 *rxbuf_end = rxbuf + n_rx;
>  	int ret = 0;
>  
> +	if (!f_pdata->use_direct_mode)
> +		writel(trigger_address, reg_base + CQSPI_REG_INDIRECTTRIGGER);
> +

Again, as I pointed out previously, this should not be needed.
cqspi_controller_init() already does above configuration and no need to
touch this register on every read.

>  	writel(from_addr, reg_base + CQSPI_REG_INDIRECTRDSTARTADDR);
>  	writel(remaining, reg_base + CQSPI_REG_INDIRECTRDBYTES);
>  
> @@ -609,6 +613,14 @@ static int cqspi_write_setup(struct spi_nor *nor)
>  	struct cqspi_st *cqspi = f_pdata->cqspi;
>  	void __iomem *reg_base = cqspi->iobase;
>  
> +	/* Disable the DMA and direct access controller */
> +	if (!f_pdata->use_direct_mode) {
> +		reg = readl(reg_base + CQSPI_REG_CONFIG);
> +		reg &= ~CQSPI_REG_CONFIG_ENB_DIR_ACC_CTRL;
> +		reg &= ~CQSPI_REG_CONFIG_DMA_MASK;
> +		writel(reg, reg_base + CQSPI_REG_CONFIG);
> +	}
> +

You did not respond to my previous comment. Why would you need to clear
CQSPI_REG_CONFIG_DMA_MASK field, if reset default is 0 anyway?

>  	/* Set opcode. */
>  	reg = nor->program_opcode << CQSPI_REG_WR_INSTR_OPCODE_LSB;
>  	writel(reg, reg_base + CQSPI_REG_WR_INSTR);
> @@ -1171,7 +1183,8 @@ static int cqspi_of_get_pdata(struct platform_device *pdev)
>  		return -ENXIO;
>  	}
>  
> -	cqspi->rclk_en = of_property_read_bool(np, "cdns,rclk-en");
> +	if (!of_device_is_compatible(np, "intel,lgm-qspi"))
> +		cqspi->rclk_en = of_property_read_bool(np, "cdns,rclk-en");
>  

If you don't want to use this property, then just drop it from your DT.
Don't override it in the driver based on compatible.

>  	return 0;
>  }
> @@ -1301,7 +1314,8 @@ static int cqspi_setup_flash(struct cqspi_st *cqspi, struct device_node *np)
>  		f_pdata->registered = true;
>  
>  		if (mtd->size <= cqspi->ahb_size) {
> -			f_pdata->use_direct_mode = true;
> +			f_pdata->use_direct_mode =
> +				!(of_device_is_compatible(np, "intel,lgm-qspi"));

Looks like, you haven't followed any of my advice. Please add a quirk
flag to disable DAC mode. Something like:

#define CQSPI_DISABLE_DAC_MODE BIT(1)

static const struct cqspi_driver_platdata intel_lgm_qspi = {
        .hwcaps_mask = CQSPI_BASE_HWCAPS_MASK,
        .quirks = CQSPI_DISABLE_DAC_MODE,
};

static const struct of_device_id cqspi_dt_ids[] = {

	...

        {
                .compatible = "intel,lgm-qspi",
                .data = &intel_lgm_qspi,
        },

	...
};


Then in cqspi_setup_flash(),

	if (mtd->size <= cqspi->ahb_size &&
		!ddata->quirks & CQSPI_DISABLE_DAC_MODE) {
		f_pdata->use_direct_mode = true;
		...
	}	


>  			dev_dbg(nor->dev, "using direct mode for %s\n",
>  				mtd->name);
>  
> @@ -1347,7 +1361,7 @@ static int cqspi_probe(struct platform_device *pdev)
>  	}
>  
>  	/* Obtain QSPI clock. */
> -	cqspi->clk = devm_clk_get(dev, NULL);
> +	cqspi->clk = devm_clk_get(dev, "qspi");

This will break DT backward compatibility. Existing DTs don't have
clock-names = "qspi". Hence, this code will error out.
What I meant in the previous mail was: if device does not have multiple
clk inputs then there is no need for clock-names and there is no need to
touch this part of code.

	cqspi->clk = devm_clk_get(dev, NULL);

This should just work fine even on your board. So drop this hunk.

>  	if (IS_ERR(cqspi->clk)) {
>  		dev_err(dev, "Cannot claim QSPI clock.\n");
>  		return PTR_ERR(cqspi->clk);
> @@ -1369,6 +1383,7 @@ static int cqspi_probe(struct platform_device *pdev)
>  		return PTR_ERR(cqspi->ahb_base);
>  	}
>  	cqspi->mmap_phys_base = (dma_addr_t)res_ahb->start;
> +	cqspi->trigger_address = res_ahb->start;

Nope, this is not needed. See:
https://elixir.bootlin.com/linux/v5.3-rc6/source/drivers/mtd/spi-nor/cadence-quadspi.c#L1168

Populate the trigger-address using cdns,trigger-address property in DT

>  	cqspi->ahb_size = resource_size(res_ahb);
>  
>  	init_completion(&cqspi->transfer_complete);
> 

-- 
Regards
Vignesh

______________________________________________________
Linux MTD discussion mailing list
http://lists.infradead.org/mailman/listinfo/linux-mtd/

  reply	other threads:[~2019-09-09  6:05 UTC|newest]

Thread overview: 8+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2019-08-27  3:58 [PATCH v2 0/3] dt-bindings: mtd: cadence-qspi:add support for Intel lgm-qspi Ramuthevar, Vadivel MuruganX
2019-08-27  3:58 ` [PATCH v2 1/3] " Ramuthevar, Vadivel MuruganX
2019-09-02 13:39   ` Rob Herring
2019-09-03  2:03     ` Ramuthevar, Vadivel MuruganX
2019-08-27  3:58 ` [PATCH v2 2/3] mtd: spi-nor: cadence-quadspi: disable DMA and DAC for Intel LGM Ramuthevar, Vadivel MuruganX
2019-09-09  6:05   ` Vignesh Raghavendra [this message]
2019-09-09  7:32     ` Ramuthevar, Vadivel MuruganX
2019-08-27  3:58 ` [PATCH v2 3/3] mtd: spi-nor: cadence-quadspi: disable the auto-poll " Ramuthevar, Vadivel MuruganX

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=2596ecd4-4ba6-ff7d-472f-1f6e664b4a97@ti.com \
    --to=vigneshr@ti.com \
    --cc=andriy.shevchenko@intel.com \
    --cc=boris.brezillon@free-electrons.com \
    --cc=cheol.yong.kim@intel.com \
    --cc=computersforpeace@gmail.com \
    --cc=cyrille.pitchen@atmel.com \
    --cc=david.oberhollenzer@sigma-star.at \
    --cc=devicetree@vger.kernel.org \
    --cc=dwmw2@infradead.org \
    --cc=jwboyer@gmail.com \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-mtd@lists.infradead.org \
    --cc=miquel.raynal@bootlin.com \
    --cc=qi-ming.wu@intel.com \
    --cc=richard@nod.at \
    --cc=tudor.ambarus@gmail.com \
    --cc=vadivel.muruganx.ramuthevar@linux.intel.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).