[PATCH v3 1/2] mtd: nand: sunxi: Replace failsafe timing cfg with calculated value
Boris Brezillon
boris.brezillon at free-electrons.com
Wed Jun 10 01:59:13 PDT 2015
Hi Roy,
On Wed, 10 Jun 2015 10:29:07 +0200
Roy Spliet <r.spliet at ultimaker.com> wrote:
> Calculates the timing cfg value once when initialising a chip, then sets
> it on chip select. Register definition documented the A83 user manual.
How about rewording the sentence this way:
"
The TIMING_CFG register was previously statically set to a magic value
(extracted from Allwinner's BSP) when initializing the NAND controller.
Now that we have more details about the TIMING_CFG register layout
(extracted from the A83 user manual) we can dynamically calculate the
appropriate value for each NAND chip and set it when selecting the
chip.
"
>
> Signed-off-by: Roy Spliet <r.spliet at ultimaker.com>
>
> V2:
> - Fix crippled comments
>
> V3:
> - Warn for invalid timings
> - Style
Almost right: the changelog should be placed after the '---' line ;-).
> ---
> drivers/mtd/nand/sunxi_nand.c | 71 ++++++++++++++++++++++++++++++++++++++++---
> 1 file changed, 66 insertions(+), 5 deletions(-)
>
> diff --git a/drivers/mtd/nand/sunxi_nand.c b/drivers/mtd/nand/sunxi_nand.c
> index 6f93b29..68090b7 100644
> --- a/drivers/mtd/nand/sunxi_nand.c
> +++ b/drivers/mtd/nand/sunxi_nand.c
> @@ -99,6 +99,13 @@
> NFC_CMD_INT_ENABLE | \
> NFC_DMA_INT_ENABLE)
>
> +/* define bit use in NFC_TIMING_CFG */
/* define NFC_TIMING_CFG register layout */
> +#define NFC_TIMING_CFG(tWB, tADL, tWHR, tRHW, tCAD) \
> + ((tWB) & 0x3) | (((tADL) & 0x3) << 2) | \
> + (((tWHR) & 0x3) << 4) | (((tRHW) & 0x3) << 6) | \
> + (((tCAD) & 0x7) << 8);
> +
> +
> /* define bit use in NFC_CMD */
> #define NFC_CMD_LOW_BYTE GENMASK(7, 0)
> #define NFC_CMD_HIGH_BYTE GENMASK(15, 8)
> @@ -208,6 +215,7 @@ struct sunxi_nand_hw_ecc {
> * @nand: base NAND chip structure
> * @mtd: base MTD structure
> * @clk_rate: clk_rate required for this NAND chip
> + * @timing_cfg TIMING_CFG register value for this NAND chip
> * @selected: current active CS
> * @nsels: number of CS lines required by the NAND chip
> * @sels: array of CS lines descriptions
> @@ -217,6 +225,7 @@ struct sunxi_nand_chip {
> struct nand_chip nand;
> struct mtd_info mtd;
> unsigned long clk_rate;
> + u32 timing_cfg;
> int selected;
> int nsels;
> struct sunxi_nand_chip_sel sels[0];
> @@ -403,6 +412,7 @@ static void sunxi_nfc_select_chip(struct mtd_info *mtd, int chip)
> }
> }
>
> + writel(sunxi_nand->timing_cfg, nfc->regs + NFC_REG_TIMING_CFG);
> writel(ctl, nfc->regs + NFC_REG_CTL);
>
> sunxi_nand->selected = chip;
> @@ -807,10 +817,33 @@ static int sunxi_nfc_hw_syndrome_ecc_write_page(struct mtd_info *mtd,
> return 0;
> }
>
> +static const s32 tWB_lut[] = {6, 12, 16, 20};
> +static const s32 tRHW_lut[] = {4, 8, 12, 20};
> +
> +static s32 _sunxi_nand_lookup_timing(const s32 *lut, int lut_size, u32 duration,
> + u32 clk_period)
You can return an int here.
> +{
> + u32 clk_cycles = (duration + clk_period - 1) / clk_period;
DIV_ROUND_UP(duration, clk_period) ??
> + int i;
> +
> + for (i = 0; i < lut_size; i++) {
> + if (clk_cycles <= lut[i])
> + return i;
> + }
> +
> + /* Doesn't fit */
> + return -1;
How about returning -EINVAL (or -ENOTSUP) directly ?
> +}
> +
> +#define sunxi_nand_lookup_timing(l,p,c) \
> + _sunxi_nand_lookup_timing(l, ARRAY_SIZE(l), p, c)
> +
> static int sunxi_nand_chip_set_timings(struct sunxi_nand_chip *chip,
> const struct nand_sdr_timings *timings)
> {
> + struct sunxi_nfc *nfc = to_sunxi_nfc(chip->nand.controller);
> u32 min_clk_period = 0;
> + u32 tWB, tADL, tWHR, tRHW, tCAD;
>
> /* T1 <=> tCLS */
> if (timings->tCLS_min > min_clk_period)
> @@ -872,6 +905,37 @@ static int sunxi_nand_chip_set_timings(struct sunxi_nand_chip *chip,
> if (timings->tWC_min > (min_clk_period * 2))
> min_clk_period = DIV_ROUND_UP(timings->tWC_min, 2);
>
> + /* T16 - T19 + tCAD */
> + tWB = sunxi_nand_lookup_timing(tWB_lut, timings->tWB_max,
> + min_clk_period);
> + if (tWB == -1) {
> + dev_err(nfc->dev, "unsupported tWB\n");
> + return -EINVAL;
If you return -EINVAL in case of failure, you can just replace that
test by
if (tWB < 0) {
dev_err(nfc->dev, "unsupported tWB\n");
return tWB;
}
> + }
> +
> + tADL = DIV_ROUND_UP(timings->tADL_min, min_clk_period) >> 3;
> + if (tADL > 3) {
> + dev_err(nfc->dev, "unsupported tADL\n");
> + return -EINVAL;
> + }
> +
> + tWHR = DIV_ROUND_UP(timings->tWHR_min, min_clk_period) >> 3;
> + if (tWHR > 3) {
> + dev_err(nfc->dev, "unsupported tWHR\n");
> + return -EINVAL;
> + }
> +
> + tRHW = sunxi_nand_lookup_timing(tRHW_lut, timings->tRHW_min,
> + min_clk_period);
> + if (tRHW == -1) {
> + dev_err(nfc->dev, "unsupported tRHW\n");
> + return -EINVAL;
> + }
Ditto
> +
> + tCAD = 0x7;
Can you add a comment explaining why you're hardcoding the tCAD value
(AFAIR, tCAD is specific to DDR NANDs, and the NAND layer does not
support DDR NAND timings yet, maybe we should also add a FIXME or TODO
tag).
> +
> + /* TODO: A83 has some more bits for CDQSS, CS, CLHZ, CCS, WC */
> + chip->timing_cfg = NFC_TIMING_CFG(tWB, tADL, tWHR, tRHW, tCAD);
>
> /* Convert min_clk_period from picoseconds to nanoseconds */
> min_clk_period = DIV_ROUND_UP(min_clk_period, 1000);
> @@ -884,8 +948,6 @@ static int sunxi_nand_chip_set_timings(struct sunxi_nand_chip *chip,
> */
> chip->clk_rate = (2 * NSEC_PER_SEC) / min_clk_period;
>
> - /* TODO: configure T16-T19 */
> -
> return 0;
> }
>
> @@ -1168,6 +1230,7 @@ static int sunxi_nand_chip_init(struct device *dev, struct sunxi_nfc *nfc,
>
> chip->nsels = nsels;
> chip->selected = -1;
> + chip->timing_cfg = 0x7ff;
Why do you need to initialize this field ?
>
> for (i = 0; i < nsels; i++) {
> ret = of_property_read_u32_index(np, "reg", i, &tmp);
> @@ -1377,11 +1440,9 @@ static int sunxi_nfc_probe(struct platform_device *pdev)
> platform_set_drvdata(pdev, nfc);
>
> /*
> - * TODO: replace these magic values with proper flags as soon as we
> - * know what they are encoding.
> + * TODO: replace this magic value with EDO flag
> */
> writel(0x100, nfc->regs + NFC_REG_TIMING_CTL);
> - writel(0x7ff, nfc->regs + NFC_REG_TIMING_CFG);
>
> ret = sunxi_nand_chips_init(dev, nfc);
> if (ret) {
--
Boris Brezillon, Free Electrons
Embedded Linux and Kernel engineering
http://free-electrons.com
More information about the linux-mtd
mailing list