Re: [PATCH] mtd: rawnand: meson: fix the clock after discarding sd_emmc_c_clkc

From: Liang Yang
Date: Sun Jan 09 2022 - 22:29:34 EST


Hi,

On 2022/1/7 17:57, Neil Armstrong wrote:
[ EXTERNAL EMAIL ]

Hi,

On 07/01/2022 05:04, Liang Yang wrote:
Hi Jerome,

On 2022/1/6 16:16, Jerome Brunet wrote:
[ EXTERNAL EMAIL ]


On Thu 06 Jan 2022 at 11:25, Liang Yang <liang.yang@xxxxxxxxxxx> wrote:

Because EMMC and NAND have the same control clock register, so we
implement a 'sd_emmc_c_clkc'. Previously DTS is defined as below:

    sd_emmc_c_clkc: mmc@7000 {
        compatible = "amlogic,meson-axg-mmc-clkc", "syscon";
        reg = <0x0 0x7000 0x0 0x800>;
    };

    nand-controller@7800 {
        ......

        clocks = <&clkc CLKID_SD_EMMC_C>,
            <&sd_emmc_c_clkc CLKID_MMC_DIV>,
            <&sd_emmc_c_clkc CLKID_MMC_PHASE_RX>,
            <&sd_emmc_c_clkc CLKID_MMC_PHASE_TX>;
        clock-names = "core", "device", "rx", "tx";
        amlogic,mmc-syscon = <&sd_emmc_c_clkc>;

        ......
    }

but in fact, above implementation is rejected. so now registering
a nand_divider.

Can you give a bit of context ? a link to the discussion rejecting this

Maybe I mistake the meaning of [jianxin.pan@xxxxxxxxxxx], so i need him to clarify it.
As far as remember, things were getting done for A1 and stopped before
clock part was finished. I'm saying the change is wrong, just that a
discussion is needed before a decision is made.

ok.
previous discussion in this link:
https://lore.kernel.org/all/86789d73-5a6c-7729-ecd1-dcd342b2fcde@xxxxxxxxxxx/
it never go on since v6.

Looking at the serie I don't see any major blockers, if you're willing to take ownership
of this patchset and send a v6 with all issues/comments addressed, it would be easier than
restarting the whole design discussion

Neil


Hi Neil, thanks for your advice.
After talking with [jianxin.pan], form this link <https://lkml.org/lkml/2019/1/8/441>, it already moves futher and on v9, so we decide to continue to work hard on it.




Change-Id: Ibeb4c7ff886f5886aac4d6c664d7bbd1b1bcb997
Signed-off-by: Liang Yang <liang.yang@xxxxxxxxxxx>
---
  drivers/mtd/nand/raw/meson_nand.c | 88 +++++++++++++++++--------------
  1 file changed, 49 insertions(+), 39 deletions(-)

diff --git a/drivers/mtd/nand/raw/meson_nand.c b/drivers/mtd/nand/raw/meson_nand.c
index ac3be92872d0..4472363059c2 100644
--- a/drivers/mtd/nand/raw/meson_nand.c
+++ b/drivers/mtd/nand/raw/meson_nand.c
@@ -2,7 +2,7 @@
  /*
   * Amlogic Meson Nand Flash Controller Driver
   *
- * Copyright (c) 2018 Amlogic, inc.
+ * Copyright (c) 2018-2021 Amlogic, inc.
   * Author: Liang Yang <liang.yang@xxxxxxxxxxx>
   */
  @@ -10,6 +10,7 @@
  #include <linux/dma-mapping.h>
  #include <linux/interrupt.h>
  #include <linux/clk.h>
+#include <linux/clk-provider.h>
  #include <linux/mtd/rawnand.h>
  #include <linux/mtd/mtd.h>
  #include <linux/mfd/syscon.h>
@@ -55,6 +56,7 @@
  #define NFC_REG_VER        0x38
    #define NFC_RB_IRQ_EN        BIT(21)
+#define NFC_CMD_FIFO_RESET    BIT(31)
    #define CMDRWGEN(cmd_dir, ran, bch, short_mode, page_size, pages)    \
      (                                \
@@ -104,6 +106,9 @@
    #define PER_INFO_BYTE        8
  +#define CLK_DIV_SHIFT        0
+#define CLK_DIV_WIDTH        6
+
  struct meson_nfc_nand_chip {
      struct list_head node;
      struct nand_chip nand;
@@ -151,15 +156,15 @@ struct meson_nfc {
      struct nand_controller controller;
      struct clk *core_clk;
      struct clk *device_clk;
-    struct clk *phase_tx;
-    struct clk *phase_rx;

I would have been nice to explain why these clock are no longer
required, in the change description maybe ?

i got the reply from our vlsi members that rx and tx doesn't make sense to meson NFC timing. so previous commit about this should be wrong.

+    struct clk *nand_clk;
+    struct clk_divider nand_divider;
        unsigned long clk_rate;
      u32 bus_timing;
        struct device *dev;
      void __iomem *reg_base;
-    struct regmap *reg_clk;
+    void __iomem *reg_clk;
      struct completion completion;
      struct list_head chips;
      const struct meson_nfc_data *data;
@@ -406,12 +411,14 @@ static int meson_nfc_queue_rb(struct meson_nfc *nfc, int timeout_ms)
      cmd = NFC_CMD_RB | NFC_CMD_RB_INT
          | nfc->param.chip_select | nfc->timing.tbers_max;
      writel(cmd, nfc->reg_base + NFC_REG_CMD);
-
+    meson_nfc_drain_cmd(nfc);

is this clock related ?
ok, i will split it in another patch.

      ret = wait_for_completion_timeout(&nfc->completion,
                        msecs_to_jiffies(timeout_ms));
      if (ret == 0)
          ret = -1;
  +    /* reset command fifo to avoid lock */
+    writel(NFC_CMD_FIFO_RESET, nfc->reg_base + NFC_REG_CMD);

Again, not seem to be clock related - does not belong in this patch
ok, i will split it in another patch.

      return ret;
  }
  @@ -988,8 +995,9 @@ static const struct mtd_ooblayout_ops meson_ooblayout_ops = {
  static int meson_nfc_clk_init(struct meson_nfc *nfc)
  {
      int ret;
+    struct clk_init_data init = {0};
+    const char *fix_div2_pll_name[1];
  -    /* request core clock */

Why is the comment bothering you ?

      nfc->core_clk = devm_clk_get(nfc->dev, "core");
      if (IS_ERR(nfc->core_clk)) {
          dev_err(nfc->dev, "failed to get core clock\n");
@@ -1002,21 +1010,25 @@ static int meson_nfc_clk_init(struct meson_nfc *nfc)
          return PTR_ERR(nfc->device_clk);
      }
  -    nfc->phase_tx = devm_clk_get(nfc->dev, "tx");
-    if (IS_ERR(nfc->phase_tx)) {
-        dev_err(nfc->dev, "failed to get TX clk\n");
-        return PTR_ERR(nfc->phase_tx);
-    }
-
-    nfc->phase_rx = devm_clk_get(nfc->dev, "rx");
-    if (IS_ERR(nfc->phase_rx)) {
-        dev_err(nfc->dev, "failed to get RX clk\n");
-        return PTR_ERR(nfc->phase_rx);
-    }
+    init.name = devm_kstrdup(nfc->dev, "nfc#div", GFP_KERNEL);
+    init.ops = &clk_divider_ops;
+    fix_div2_pll_name[0] = __clk_get_name(nfc->device_clk);

You should be using "fw_name" so CCF looks a DT directly, instead of this
ok

+    init.parent_names = fix_div2_pll_name;
+    init.num_parents = 1;
+    nfc->nand_divider.reg = nfc->reg_clk;
+    nfc->nand_divider.shift = CLK_DIV_SHIFT;
+    nfc->nand_divider.width = CLK_DIV_WIDTH;
+    nfc->nand_divider.hw.init = &init;
+    nfc->nand_divider.flags = CLK_DIVIDER_ONE_BASED |
+                  CLK_DIVIDER_ROUND_CLOSEST |
+                        CLK_DIVIDER_ALLOW_ZERO;
+
+    nfc->nand_clk = devm_clk_register(nfc->dev, &nfc->nand_divider.hw);
+    if (IS_ERR(nfc->nand_clk))
+        return PTR_ERR(nfc->nand_clk);
        /* init SD_EMMC_CLOCK to sane defaults w/min clock rate */
-    regmap_update_bits(nfc->reg_clk,
-               0, CLK_SELECT_NAND, CLK_SELECT_NAND);
+    writel(CLK_SELECT_NAND | readl(nfc->reg_clk), nfc->reg_clk);
        ret = clk_prepare_enable(nfc->core_clk);
      if (ret) {
@@ -1030,29 +1042,21 @@ static int meson_nfc_clk_init(struct meson_nfc *nfc)
          goto err_device_clk;
      }
  -    ret = clk_prepare_enable(nfc->phase_tx);
+    ret = clk_prepare_enable(nfc->nand_clk);
      if (ret) {
-        dev_err(nfc->dev, "failed to enable TX clock\n");
-        goto err_phase_tx;
-    }
-
-    ret = clk_prepare_enable(nfc->phase_rx);
-    if (ret) {
-        dev_err(nfc->dev, "failed to enable RX clock\n");
-        goto err_phase_rx;
+        dev_err(nfc->dev, "pre enable NFC divider fail\n");
+        goto err_nand_clk;
      }
        ret = clk_set_rate(nfc->device_clk, 24000000);
      if (ret)
-        goto err_disable_rx;
+        goto err_disable_clk;
        return 0;
  -err_disable_rx:
-    clk_disable_unprepare(nfc->phase_rx);
-err_phase_rx:
-    clk_disable_unprepare(nfc->phase_tx);
-err_phase_tx:
+err_disable_clk:
+    clk_disable_unprepare(nfc->nand_clk);
+err_nand_clk:
      clk_disable_unprepare(nfc->device_clk);
  err_device_clk:
      clk_disable_unprepare(nfc->core_clk);
@@ -1061,8 +1065,7 @@ static int meson_nfc_clk_init(struct meson_nfc *nfc)
    static void meson_nfc_disable_clk(struct meson_nfc *nfc)
  {
-    clk_disable_unprepare(nfc->phase_rx);
-    clk_disable_unprepare(nfc->phase_tx);
+    clk_disable_unprepare(nfc->nand_clk);
      clk_disable_unprepare(nfc->device_clk);
      clk_disable_unprepare(nfc->core_clk);
  }
@@ -1375,6 +1378,7 @@ static int meson_nfc_probe(struct platform_device *pdev)
      struct device *dev = &pdev->dev;
      struct meson_nfc *nfc;
      struct resource *res;
+    u32 ext_clk_reg;
      int ret, irq;
        nfc = devm_kzalloc(dev, sizeof(*nfc), GFP_KERNEL);
@@ -1396,9 +1400,15 @@ static int meson_nfc_probe(struct platform_device *pdev)
      if (IS_ERR(nfc->reg_base))
          return PTR_ERR(nfc->reg_base);
  -    nfc->reg_clk =
-        syscon_regmap_lookup_by_phandle(dev->of_node,
-                        "amlogic,mmc-syscon");
+    ret = of_property_read_u32(pdev->dev.of_node,
+                   "sd_emmc_c_clkc",
+                   &ext_clk_reg);
+    if (ret) {
+        dev_err(dev, "failed to get NAND external clock register\n");
+        return ret;
+    }
+
+    nfc->reg_clk = devm_ioremap(&pdev->dev, ext_clk_reg, sizeof(int));

That's how you should get a register region.
If you want an example of a device claiming several region on amlogic,
have a look at the pinctrl driver.
ok

      if (IS_ERR(nfc->reg_clk)) {
          dev_err(dev, "Failed to lookup clock base\n");
          return PTR_ERR(nfc->reg_clk);

.

.