Re: [PATCH v5 2/2] mtd: spi-nor: cadence-quadspi: Add support for Octal SPI controller

From: Vignesh R
Date: Tue Feb 05 2019 - 01:12:56 EST


Hi,

On 29/01/19 9:02 PM, Tudor.Ambarus@xxxxxxxxxxxxx wrote:
>
>
> On 01/28/2019 07:49 AM, Vignesh R wrote:
>> Cadence OSPI controller IP supports Octal IO (x8 IO lines),
>> It also has an integrated PHY. IP register layout is very
>> similar to existing QSPI IP except for additional bits to support Octal
>> and Octal DDR mode. Therefore, extend current driver to support Octal
>> mode. Only Octal SDR read (1-1-8)mode is supported for now.
>>
>> Tested with mt35xu512aba Octal flash on TI's AM654 EVM.
>>
>> Signed-off-by: Vignesh R <vigneshr@xxxxxx>
>> Reviewed-by: Tudor Ambarus <tudor.ambarus@xxxxxxxxxxxxx>
>> ---
>>
>> v5:
>> s/cqsi_base_hwcaps_mask/CQSPI_BASE_HWCAPS_MASK/g
>> Add back cqspi_driver_platdata definition for base compatible.
>>
>> v4: Fix comments by Tudor on v3
>> v3: No changes
>> v2: Declare Octal mode capability based on compatible.
>>
>> drivers/mtd/spi-nor/cadence-quadspi.c | 58 +++++++++++++++++++++------
>> 1 file changed, 46 insertions(+), 12 deletions(-)
[...]
>>
>> static int cqspi_setup_flash(struct cqspi_st *cqspi, struct device_node *np)
>> {
>> - const struct spi_nor_hwcaps hwcaps = {
>> - .mask = SNOR_HWCAPS_READ |
>> - SNOR_HWCAPS_READ_FAST |
>> - SNOR_HWCAPS_READ_1_1_2 |
>> - SNOR_HWCAPS_READ_1_1_4 |
>> - SNOR_HWCAPS_PP,
>> - };
>> struct platform_device *pdev = cqspi->pdev;
>> struct device *dev = &pdev->dev;
>> + const struct cqspi_driver_platdata *ddata;
>> + struct spi_nor_hwcaps hwcaps;
>> struct cqspi_flash_pdata *f_pdata;
>> struct spi_nor *nor;
>> struct mtd_info *mtd;
>> unsigned int cs;
>> int i, ret;
>>
>> + ddata = of_device_get_match_data(dev);
>> + if (!ddata)
>> + hwcaps.mask = CQSPI_BASE_HWCAPS_MASK;
>
> Now that .data is set in all cqspi_dt_ids[], maybe it's better to print a
> message and return an error here. But I guess it's a matter of taste, so not a
> show stopper.

Since, driver data is kernel internal field, I guess there is little
help in printing out the error to the user when its missing. I prefer to
keep this as is, as basic Quad mode is supported by all versions of the IP.

Regards
Vignesh

>
>> + else
>> + hwcaps.mask = ddata->hwcaps_mask;
>> +
>> /* Get flash device data */
>> for_each_available_child_of_node(dev->of_node, np) {
>> ret = of_property_read_u32(np, "reg", &cs);
>> @@ -1310,7 +1326,7 @@ static int cqspi_probe(struct platform_device *pdev)
>> struct cqspi_st *cqspi;
>> struct resource *res;
>> struct resource *res_ahb;
>> - unsigned long data;
>> + const struct cqspi_driver_platdata *ddata;
>> int ret;
>> int irq;
>>
>> @@ -1377,8 +1393,8 @@ static int cqspi_probe(struct platform_device *pdev)
>> }
>>
>> cqspi->master_ref_clk_hz = clk_get_rate(cqspi->clk);
>> - data = (unsigned long)of_device_get_match_data(dev);
>> - if (data & CQSPI_NEEDS_WR_DELAY)
>> + ddata = of_device_get_match_data(dev);
>> + if (ddata && (ddata->quirks & CQSPI_NEEDS_WR_DELAY))
>> cqspi->wr_delay = 5 * DIV_ROUND_UP(NSEC_PER_SEC,
>> cqspi->master_ref_clk_hz);
>>
>> @@ -1460,14 +1476,32 @@ static const struct dev_pm_ops cqspi__dev_pm_ops = {
>> #define CQSPI_DEV_PM_OPS NULL
>> #endif
>>
>> +static const struct cqspi_driver_platdata cdns_qspi = {
>> + .hwcaps_mask = CQSPI_BASE_HWCAPS_MASK,
>> +};
>> +
>> +static const struct cqspi_driver_platdata k2g_qspi = {
>> + .hwcaps_mask = CQSPI_BASE_HWCAPS_MASK,
>> + .quirks = CQSPI_NEEDS_WR_DELAY,
>> +};
>> +
>> +static const struct cqspi_driver_platdata am654_ospi = {
>> + .hwcaps_mask = CQSPI_BASE_HWCAPS_MASK | SNOR_HWCAPS_READ_1_1_8,
>> + .quirks = CQSPI_NEEDS_WR_DELAY,
>> +};
>> +
>> static const struct of_device_id cqspi_dt_ids[] = {
>> {
>> .compatible = "cdns,qspi-nor",
>> - .data = (void *)0,
>> + .data = &cdns_qspi,
>> },
>> {
>> .compatible = "ti,k2g-qspi",
>> - .data = (void *)CQSPI_NEEDS_WR_DELAY,
>> + .data = &k2g_qspi,
>> + },
>> + {
>> + .compatible = "ti,am654-ospi",
>> + .data = &am654_ospi,
>> },
>> { /* end of table */ }
>> };
>>

--
Regards
Vignesh