[PATCH v5 0/2] mmc: sdhci-of-dwcmshc: Add CQE support

[PATCH v5 0/2] mmc: sdhci-of-dwcmshc: Add CQE support

[Sergey Khimich]

Hello!

This is implementation of SDHCI CQE support for sdhci-of-dwcmshc driver.
For enabling CQE support just set 'supports-cqe' in your DevTree file
for appropriate mmc node.

Also, while implementing CQE support for the driver, I faced with a problem
which I will describe below.
According to the IP block documentation CQE works only with "AMDA-2 only"
mode which is activated only with v4 mode enabled. I see in dwcmshc_probe()
function that v4 mode gets enabled only for 'sdhci_dwcmshc_bf3_pdata'
platform data.

So my question is: is it correct to enable v4 mode for all platform data
if 'SDHCI_CAN_64BIT_V4' bit is set in hw?

Because I`m afraid that enabling v4 mode for some platforms could break
them down. On the other hand, if host controller says that it can do v4
(caps & SDHCI_CAN_64BIT_V4), lets do v4 or disable it manualy by some
quirk. Anyway - RFC.


v2:
 - Added dwcmshc specific cqe_disable hook to prevent losing
   in-flight cmd when an ioctl is issued and cqe_disable is called;

 - Added processing 128Mb boundary for the host memory data buffer size
   and the data buffer. For implementing this processing an extra
   callback is added to the struct 'sdhci_ops'.

 - Fixed typo.

v3:
 - Fix warning reported by kernel test robot:
        | Reported-by: kernel test robot <lkp@intel.com>
        | Closes: https://lore.kernel.org/oe-kbuild-all/202309270807.VoVn81m6-lkp@intel.com/
        | Closes: https://lore.kernel.org/oe-kbuild-all/202309300806.dcR19kcE-lkp@intel.com/

v4:
 - Data reset moved to custom driver tuning hook.
 - Removed unnecessary dwcmshc_sdhci_cqe_disable() func
 - Removed unnecessary dwcmshc_cqhci_set_tran_desc. Export and use
   cqhci_set_tran_desc() instead.
 - Provide a hook for cqhci_set_tran_desc() instead of cqhci_prep_tran_desc().
 - Fix typo: int_clok_disable --> int_clock_disable

v5:
 - Fix warning reported by kernel test robot:
	| Reported-by: kernel test robot <lkp@intel.com>
	| Closes: https://lore.kernel.org/oe-kbuild-all/202312301130.itEZhhI5-lkp@intel.com/




Sergey Khimich (2):
  mmc: cqhci: Add cqhci set_tran_desc() callback
  mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

 drivers/mmc/host/Kconfig            |   1 +
 drivers/mmc/host/cqhci-core.c       |  10 +-
 drivers/mmc/host/cqhci.h            |   5 +
 drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
 4 files changed, 192 insertions(+), 5 deletions(-)

-- 
2.30.2

[PATCH v5 1/2] mmc: cqhci: Add cqhci set_tran_desc() callback

[Sergey Khimich]

From: Sergey Khimich <serghox@gmail.com>

There are could be specific limitations for some mmc
controllers for setting cqhci transfer descriptors.
So add callback to allow implement driver specific function.

Signed-off-by: Sergey Khimich <serghox@gmail.com>
---
 drivers/mmc/host/cqhci-core.c | 10 +++++++---
 drivers/mmc/host/cqhci.h      |  5 +++++
 2 files changed, 12 insertions(+), 3 deletions(-)

diff --git a/drivers/mmc/host/cqhci-core.c b/drivers/mmc/host/cqhci-core.c
index 41e94cd14109..d12870b124cc 100644
--- a/drivers/mmc/host/cqhci-core.c
+++ b/drivers/mmc/host/cqhci-core.c
@@ -474,8 +474,8 @@ static int cqhci_dma_map(struct mmc_host *host, struct mmc_request *mrq)
 	return sg_count;
 }
 
-static void cqhci_set_tran_desc(u8 *desc, dma_addr_t addr, int len, bool end,
-				bool dma64)
+void cqhci_set_tran_desc(u8 *desc, dma_addr_t addr, int len, bool end,
+			 bool dma64)
 {
 	__le32 *attr = (__le32 __force *)desc;
 
@@ -522,7 +522,11 @@ static int cqhci_prep_tran_desc(struct mmc_request *mrq,
 
 		if ((i+1) == sg_count)
 			end = true;
-		cqhci_set_tran_desc(desc, addr, len, end, dma64);
+		if (cq_host->ops->set_tran_desc)
+			cq_host->ops->set_tran_desc(cq_host, &desc, addr, len, end, dma64);
+		else
+			cqhci_set_tran_desc(desc, addr, len, end, dma64);
+
 		desc += cq_host->trans_desc_len;
 	}
 
diff --git a/drivers/mmc/host/cqhci.h b/drivers/mmc/host/cqhci.h
index 1a12e40a02e6..703d5af6c49e 100644
--- a/drivers/mmc/host/cqhci.h
+++ b/drivers/mmc/host/cqhci.h
@@ -217,6 +217,7 @@ struct cqhci_host_ops;
 struct mmc_host;
 struct mmc_request;
 struct cqhci_slot;
+struct mmc_data;
 
 struct cqhci_host {
 	const struct cqhci_host_ops *ops;
@@ -293,6 +294,9 @@ struct cqhci_host_ops {
 	int (*program_key)(struct cqhci_host *cq_host,
 			   const union cqhci_crypto_cfg_entry *cfg, int slot);
 #endif
+	void (*set_tran_desc)(struct cqhci_host *cq_host, u8 **desc,
+			      dma_addr_t addr, int len, bool end, bool dma64);
+
 };
 
 static inline void cqhci_writel(struct cqhci_host *host, u32 val, int reg)
@@ -318,6 +322,7 @@ irqreturn_t cqhci_irq(struct mmc_host *mmc, u32 intmask, int cmd_error,
 int cqhci_init(struct cqhci_host *cq_host, struct mmc_host *mmc, bool dma64);
 struct cqhci_host *cqhci_pltfm_init(struct platform_device *pdev);
 int cqhci_deactivate(struct mmc_host *mmc);
+void cqhci_set_tran_desc(u8 *desc, dma_addr_t addr, int len, bool end, bool dma64);
 static inline int cqhci_suspend(struct mmc_host *mmc)
 {
 	return cqhci_deactivate(mmc);
-- 
2.30.2

Re: [PATCH v5 1/2] mmc: cqhci: Add cqhci set_tran_desc() callback

[Adrian Hunter]

On 31/12/23 16:46, Sergey Khimich wrote:
> From: Sergey Khimich <serghox@gmail.com>
> 
> There are could be specific limitations for some mmc
> controllers for setting cqhci transfer descriptors.
> So add callback to allow implement driver specific function.
> 
> Signed-off-by: Sergey Khimich <serghox@gmail.com>
> ---
>  drivers/mmc/host/cqhci-core.c | 10 +++++++---
>  drivers/mmc/host/cqhci.h      |  5 +++++
>  2 files changed, 12 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/mmc/host/cqhci-core.c b/drivers/mmc/host/cqhci-core.c
> index 41e94cd14109..d12870b124cc 100644
> --- a/drivers/mmc/host/cqhci-core.c
> +++ b/drivers/mmc/host/cqhci-core.c
> @@ -474,8 +474,8 @@ static int cqhci_dma_map(struct mmc_host *host, struct mmc_request *mrq)
>  	return sg_count;
>  }
>  
> -static void cqhci_set_tran_desc(u8 *desc, dma_addr_t addr, int len, bool end,
> -				bool dma64)
> +void cqhci_set_tran_desc(u8 *desc, dma_addr_t addr, int len, bool end,

Needs to be exported also i.e.
EXPORT_SYMBOL(cqhci_set_tran_desc);

> +			 bool dma64)
>  {
>  	__le32 *attr = (__le32 __force *)desc;
>  
> @@ -522,7 +522,11 @@ static int cqhci_prep_tran_desc(struct mmc_request *mrq,
>  
>  		if ((i+1) == sg_count)
>  			end = true;
> -		cqhci_set_tran_desc(desc, addr, len, end, dma64);
> +		if (cq_host->ops->set_tran_desc)
> +			cq_host->ops->set_tran_desc(cq_host, &desc, addr, len, end, dma64);
> +		else
> +			cqhci_set_tran_desc(desc, addr, len, end, dma64);
> +
>  		desc += cq_host->trans_desc_len;
>  	}
>  
> diff --git a/drivers/mmc/host/cqhci.h b/drivers/mmc/host/cqhci.h
> index 1a12e40a02e6..703d5af6c49e 100644
> --- a/drivers/mmc/host/cqhci.h
> +++ b/drivers/mmc/host/cqhci.h
> @@ -217,6 +217,7 @@ struct cqhci_host_ops;
>  struct mmc_host;
>  struct mmc_request;
>  struct cqhci_slot;
> +struct mmc_data;

Not used

>  
>  struct cqhci_host {
>  	const struct cqhci_host_ops *ops;
> @@ -293,6 +294,9 @@ struct cqhci_host_ops {
>  	int (*program_key)(struct cqhci_host *cq_host,
>  			   const union cqhci_crypto_cfg_entry *cfg, int slot);
>  #endif
> +	void (*set_tran_desc)(struct cqhci_host *cq_host, u8 **desc,
> +			      dma_addr_t addr, int len, bool end, bool dma64);
> +
>  };
>  
>  static inline void cqhci_writel(struct cqhci_host *host, u32 val, int reg)
> @@ -318,6 +322,7 @@ irqreturn_t cqhci_irq(struct mmc_host *mmc, u32 intmask, int cmd_error,
>  int cqhci_init(struct cqhci_host *cq_host, struct mmc_host *mmc, bool dma64);
>  struct cqhci_host *cqhci_pltfm_init(struct platform_device *pdev);
>  int cqhci_deactivate(struct mmc_host *mmc);
> +void cqhci_set_tran_desc(u8 *desc, dma_addr_t addr, int len, bool end, bool dma64);
>  static inline int cqhci_suspend(struct mmc_host *mmc)
>  {
>  	return cqhci_deactivate(mmc);

[PATCH v5 2/2] mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

[Sergey Khimich]

From: Sergey Khimich <serghox@gmail.com>

For enabling CQE support just set 'supports-cqe' in your DevTree file
for appropriate mmc node.

Signed-off-by: Sergey Khimich <serghox@gmail.com>
---
 drivers/mmc/host/Kconfig            |   1 +
 drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
 2 files changed, 180 insertions(+), 2 deletions(-)

diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
index 58bd5fe4cd25..f7594705b013 100644
--- a/drivers/mmc/host/Kconfig
+++ b/drivers/mmc/host/Kconfig
@@ -233,6 +233,7 @@ config MMC_SDHCI_OF_DWCMSHC
 	depends on MMC_SDHCI_PLTFM
 	depends on OF
 	depends on COMMON_CLK
+	select MMC_CQHCI
 	help
 	  This selects Synopsys DesignWare Cores Mobile Storage Controller
 	  support.
diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
index 3a3bae6948a8..0ba1df4bcf36 100644
--- a/drivers/mmc/host/sdhci-of-dwcmshc.c
+++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
@@ -20,6 +20,7 @@
 #include <linux/sizes.h>
 
 #include "sdhci-pltfm.h"
+#include "cqhci.h"
 
 #define SDHCI_DWCMSHC_ARG2_STUFF	GENMASK(31, 16)
 
@@ -36,6 +37,9 @@
 #define DWCMSHC_ENHANCED_STROBE		BIT(8)
 #define DWCMSHC_EMMC_ATCTRL		0x40
 
+/* DWC IP vendor area 2 pointer */
+#define DWCMSHC_P_VENDOR_AREA2		0xea
+
 /* Rockchip specific Registers */
 #define DWCMSHC_EMMC_DLL_CTRL		0x800
 #define DWCMSHC_EMMC_DLL_RXCLK		0x804
@@ -75,6 +79,11 @@
 #define BOUNDARY_OK(addr, len) \
 	((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
 
+#define DWCMSHC_SDHCI_CQE_TRNS_MODE	(SDHCI_TRNS_MULTI | \
+					 SDHCI_TRNS_BLK_CNT_EN | \
+					 SDHCI_TRNS_DMA)
+
+
 enum dwcmshc_rk_type {
 	DWCMSHC_RK3568,
 	DWCMSHC_RK3588,
@@ -90,7 +99,9 @@ struct rk35xx_priv {
 
 struct dwcmshc_priv {
 	struct clk	*bus_clk;
-	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA reg */
+	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA1 reg */
+	int vendor_specific_area2; /* P_VENDOR_SPECIFIC_AREA2 reg */
+
 	void *priv; /* pointer to SoC private stuff */
 };
 
@@ -210,6 +221,90 @@ static void dwcmshc_hs400_enhanced_strobe(struct mmc_host *mmc,
 	sdhci_writel(host, vendor, reg);
 }
 
+static int dwcmshc_execute_tuning(struct mmc_host *mmc, u32 opcode)
+{
+	int err = sdhci_execute_tuning(mmc, opcode);
+	struct sdhci_host *host = mmc_priv(mmc);
+
+	if (err)
+		return err;
+
+	/*
+	 * Tuning can leave the IP in an active state (Buffer Read Enable bit
+	 * set) which prevents the entry to low power states (i.e. S0i3). Data
+	 * reset will clear it.
+	 */
+	sdhci_reset(host, SDHCI_RESET_DATA);
+
+	return 0;
+}
+
+static u32 dwcmshc_cqe_irq_handler(struct sdhci_host *host, u32 intmask)
+{
+	int cmd_error = 0;
+	int data_error = 0;
+
+	if (!sdhci_cqe_irq(host, intmask, &cmd_error, &data_error))
+		return intmask;
+
+	cqhci_irq(host->mmc, intmask, cmd_error, data_error);
+
+	return 0;
+}
+
+static void dwcmshc_sdhci_cqe_enable(struct mmc_host *mmc)
+{
+	struct sdhci_host *host = mmc_priv(mmc);
+	u8 ctrl;
+
+	sdhci_writew(host, DWCMSHC_SDHCI_CQE_TRNS_MODE, SDHCI_TRANSFER_MODE);
+
+	sdhci_cqe_enable(mmc);
+
+	/*
+	 * The "DesignWare Cores Mobile Storage Host Controller
+	 * DWC_mshc / DWC_mshc_lite Databook" says:
+	 * when Host Version 4 Enable" is 1 in Host Control 2 register,
+	 * SDHCI_CTRL_ADMA32 bit means ADMA2 is selected.
+	 * Selection of 32-bit/64-bit System Addressing:
+	 * either 32-bit or 64-bit system addressing is selected by
+	 * 64-bit Addressing bit in Host Control 2 register.
+	 *
+	 * On the other hand the "DesignWare Cores Mobile Storage Host
+	 * Controller DWC_mshc / DWC_mshc_lite User Guide" says, that we have to
+	 * set DMA_SEL to ADMA2 _only_ mode in the Host Control 2 register.
+	 */
+	ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
+	ctrl &= ~SDHCI_CTRL_DMA_MASK;
+	ctrl |= SDHCI_CTRL_ADMA32;
+	sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
+}
+
+static void dwcmshc_set_tran_desc(struct cqhci_host *cq_host, u8 **desc,
+				  dma_addr_t addr, int len, bool end, bool dma64)
+{
+	int tmplen, offset;
+
+	if (likely(!len || BOUNDARY_OK(addr, len))) {
+		cqhci_set_tran_desc(*desc, addr, len, end, dma64);
+		return;
+	}
+
+	offset = addr & (SZ_128M - 1);
+	tmplen = SZ_128M - offset;
+	cqhci_set_tran_desc(*desc, addr, tmplen, false, dma64);
+
+	addr += tmplen;
+	len -= tmplen;
+	*desc += cq_host->trans_desc_len;
+	cqhci_set_tran_desc(*desc, addr, len, end, dma64);
+}
+
+static void dwcmshc_cqhci_dumpregs(struct mmc_host *mmc)
+{
+	sdhci_dumpregs(mmc_priv(mmc));
+}
+
 static void dwcmshc_rk3568_set_clock(struct sdhci_host *host, unsigned int clock)
 {
 	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
@@ -345,6 +440,7 @@ static const struct sdhci_ops sdhci_dwcmshc_ops = {
 	.get_max_clock		= dwcmshc_get_max_clock,
 	.reset			= sdhci_reset,
 	.adma_write_desc	= dwcmshc_adma_write_desc,
+	.irq			= dwcmshc_cqe_irq_handler,
 };
 
 static const struct sdhci_ops sdhci_dwcmshc_rk35xx_ops = {
@@ -379,6 +475,71 @@ static const struct sdhci_pltfm_data sdhci_dwcmshc_rk35xx_pdata = {
 		   SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
 };
 
+static const struct cqhci_host_ops dwcmshc_cqhci_ops = {
+	.enable		= dwcmshc_sdhci_cqe_enable,
+	.disable	= sdhci_cqe_disable,
+	.dumpregs	= dwcmshc_cqhci_dumpregs,
+	.set_tran_desc	= dwcmshc_set_tran_desc,
+};
+
+static void dwcmshc_cqhci_init(struct sdhci_host *host, struct platform_device *pdev)
+{
+	struct cqhci_host *cq_host;
+	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
+	struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
+	bool dma64 = false;
+	u16 clk;
+	int err;
+
+	host->mmc->caps2 |= MMC_CAP2_CQE | MMC_CAP2_CQE_DCMD;
+	cq_host = devm_kzalloc(&pdev->dev, sizeof(*cq_host), GFP_KERNEL);
+	if (!cq_host) {
+		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: not enough memory\n");
+		return;
+	}
+
+	/*
+	 * For dwcmshc host controller we have to enable internal clock
+	 * before access to some registers from Vendor Specific Aria 2.
+	 */
+	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
+	clk |= SDHCI_CLOCK_INT_EN;
+	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
+	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
+	if (!(clk & SDHCI_CLOCK_INT_EN)) {
+		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: internal clock enable error\n");
+		goto free_cq_host;
+	}
+
+	cq_host->mmio = host->ioaddr + priv->vendor_specific_area2;
+	cq_host->ops = &dwcmshc_cqhci_ops;
+
+	/* Enable using of 128-bit task descriptors */
+	dma64 = host->flags & SDHCI_USE_64_BIT_DMA;
+	if (dma64) {
+		dev_dbg(mmc_dev(host->mmc), "128-bit task descriptors\n");
+		cq_host->caps |= CQHCI_TASK_DESC_SZ_128;
+	}
+	err = cqhci_init(cq_host, host->mmc, dma64);
+	if (err) {
+		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: error %d\n", err);
+		goto int_clock_disable;
+	}
+
+	dev_dbg(mmc_dev(host->mmc), "CQE init done\n");
+
+	return;
+
+int_clock_disable:
+	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
+	clk &= ~SDHCI_CLOCK_INT_EN;
+	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
+
+free_cq_host:
+	devm_kfree(&pdev->dev, cq_host);
+}
+
+
 static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct dwcmshc_priv *dwc_priv)
 {
 	int err;
@@ -471,7 +632,7 @@ static int dwcmshc_probe(struct platform_device *pdev)
 	struct rk35xx_priv *rk_priv = NULL;
 	const struct sdhci_pltfm_data *pltfm_data;
 	int err;
-	u32 extra;
+	u32 extra, caps;
 
 	pltfm_data = device_get_match_data(&pdev->dev);
 	if (!pltfm_data) {
@@ -519,9 +680,12 @@ static int dwcmshc_probe(struct platform_device *pdev)
 
 	priv->vendor_specific_area1 =
 		sdhci_readl(host, DWCMSHC_P_VENDOR_AREA1) & DWCMSHC_AREA1_MASK;
+	priv->vendor_specific_area2 =
+		sdhci_readw(host, DWCMSHC_P_VENDOR_AREA2);
 
 	host->mmc_host_ops.request = dwcmshc_request;
 	host->mmc_host_ops.hs400_enhanced_strobe = dwcmshc_hs400_enhanced_strobe;
+	host->mmc_host_ops.execute_tuning = dwcmshc_execute_tuning;
 
 	if (pltfm_data == &sdhci_dwcmshc_rk35xx_pdata) {
 		rk_priv = devm_kzalloc(&pdev->dev, sizeof(struct rk35xx_priv), GFP_KERNEL);
@@ -547,6 +711,10 @@ static int dwcmshc_probe(struct platform_device *pdev)
 		sdhci_enable_v4_mode(host);
 #endif
 
+	caps = sdhci_readl(host, SDHCI_CAPABILITIES);
+	if (caps & SDHCI_CAN_64BIT_V4)
+		sdhci_enable_v4_mode(host);
+
 	host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
 
 	pm_runtime_get_noresume(dev);
@@ -557,6 +725,15 @@ static int dwcmshc_probe(struct platform_device *pdev)
 	if (err)
 		goto err_rpm;
 
+	/* Setup Command Queue Engine if enabled */
+	if (device_property_read_bool(&pdev->dev, "supports-cqe")) {
+		if (caps & SDHCI_CAN_64BIT_V4)
+			dwcmshc_cqhci_init(host, pdev);
+		else
+			dev_warn(dev, "Cannot enable CQE without V4 mode support\n");
+	}
+
+
 	if (rk_priv)
 		dwcmshc_rk35xx_postinit(host, priv);
 
-- 
2.30.2

Re: [PATCH v5 2/2] mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

[Adrian Hunter]

On 31/12/23 16:46, Sergey Khimich wrote:
> From: Sergey Khimich <serghox@gmail.com>
> 
> For enabling CQE support just set 'supports-cqe' in your DevTree file
> for appropriate mmc node.
> 
> Signed-off-by: Sergey Khimich <serghox@gmail.com>
> ---
>  drivers/mmc/host/Kconfig            |   1 +
>  drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
>  2 files changed, 180 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
> index 58bd5fe4cd25..f7594705b013 100644
> --- a/drivers/mmc/host/Kconfig
> +++ b/drivers/mmc/host/Kconfig
> @@ -233,6 +233,7 @@ config MMC_SDHCI_OF_DWCMSHC
>  	depends on MMC_SDHCI_PLTFM
>  	depends on OF
>  	depends on COMMON_CLK
> +	select MMC_CQHCI
>  	help
>  	  This selects Synopsys DesignWare Cores Mobile Storage Controller
>  	  support.
> diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
> index 3a3bae6948a8..0ba1df4bcf36 100644
> --- a/drivers/mmc/host/sdhci-of-dwcmshc.c
> +++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
> @@ -20,6 +20,7 @@
>  #include <linux/sizes.h>
>  
>  #include "sdhci-pltfm.h"
> +#include "cqhci.h"
>  
>  #define SDHCI_DWCMSHC_ARG2_STUFF	GENMASK(31, 16)
>  
> @@ -36,6 +37,9 @@
>  #define DWCMSHC_ENHANCED_STROBE		BIT(8)
>  #define DWCMSHC_EMMC_ATCTRL		0x40
>  
> +/* DWC IP vendor area 2 pointer */
> +#define DWCMSHC_P_VENDOR_AREA2		0xea
> +
>  /* Rockchip specific Registers */
>  #define DWCMSHC_EMMC_DLL_CTRL		0x800
>  #define DWCMSHC_EMMC_DLL_RXCLK		0x804
> @@ -75,6 +79,11 @@
>  #define BOUNDARY_OK(addr, len) \
>  	((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
>  
> +#define DWCMSHC_SDHCI_CQE_TRNS_MODE	(SDHCI_TRNS_MULTI | \
> +					 SDHCI_TRNS_BLK_CNT_EN | \
> +					 SDHCI_TRNS_DMA)
> +
> +
>  enum dwcmshc_rk_type {
>  	DWCMSHC_RK3568,
>  	DWCMSHC_RK3588,
> @@ -90,7 +99,9 @@ struct rk35xx_priv {
>  
>  struct dwcmshc_priv {
>  	struct clk	*bus_clk;
> -	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA reg */
> +	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA1 reg */
> +	int vendor_specific_area2; /* P_VENDOR_SPECIFIC_AREA2 reg */
> +
>  	void *priv; /* pointer to SoC private stuff */
>  };
>  
> @@ -210,6 +221,90 @@ static void dwcmshc_hs400_enhanced_strobe(struct mmc_host *mmc,
>  	sdhci_writel(host, vendor, reg);
>  }
>  
> +static int dwcmshc_execute_tuning(struct mmc_host *mmc, u32 opcode)
> +{
> +	int err = sdhci_execute_tuning(mmc, opcode);
> +	struct sdhci_host *host = mmc_priv(mmc);
> +
> +	if (err)
> +		return err;
> +
> +	/*
> +	 * Tuning can leave the IP in an active state (Buffer Read Enable bit
> +	 * set) which prevents the entry to low power states (i.e. S0i3). Data
> +	 * reset will clear it.
> +	 */
> +	sdhci_reset(host, SDHCI_RESET_DATA);
> +
> +	return 0;
> +}
> +
> +static u32 dwcmshc_cqe_irq_handler(struct sdhci_host *host, u32 intmask)
> +{
> +	int cmd_error = 0;
> +	int data_error = 0;
> +
> +	if (!sdhci_cqe_irq(host, intmask, &cmd_error, &data_error))
> +		return intmask;
> +
> +	cqhci_irq(host->mmc, intmask, cmd_error, data_error);
> +
> +	return 0;
> +}
> +
> +static void dwcmshc_sdhci_cqe_enable(struct mmc_host *mmc)
> +{
> +	struct sdhci_host *host = mmc_priv(mmc);
> +	u8 ctrl;
> +
> +	sdhci_writew(host, DWCMSHC_SDHCI_CQE_TRNS_MODE, SDHCI_TRANSFER_MODE);
> +
> +	sdhci_cqe_enable(mmc);
> +
> +	/*
> +	 * The "DesignWare Cores Mobile Storage Host Controller
> +	 * DWC_mshc / DWC_mshc_lite Databook" says:
> +	 * when Host Version 4 Enable" is 1 in Host Control 2 register,
> +	 * SDHCI_CTRL_ADMA32 bit means ADMA2 is selected.
> +	 * Selection of 32-bit/64-bit System Addressing:
> +	 * either 32-bit or 64-bit system addressing is selected by
> +	 * 64-bit Addressing bit in Host Control 2 register.
> +	 *
> +	 * On the other hand the "DesignWare Cores Mobile Storage Host
> +	 * Controller DWC_mshc / DWC_mshc_lite User Guide" says, that we have to
> +	 * set DMA_SEL to ADMA2 _only_ mode in the Host Control 2 register.
> +	 */
> +	ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
> +	ctrl &= ~SDHCI_CTRL_DMA_MASK;
> +	ctrl |= SDHCI_CTRL_ADMA32;
> +	sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
> +}
> +
> +static void dwcmshc_set_tran_desc(struct cqhci_host *cq_host, u8 **desc,
> +				  dma_addr_t addr, int len, bool end, bool dma64)
> +{
> +	int tmplen, offset;
> +
> +	if (likely(!len || BOUNDARY_OK(addr, len))) {
> +		cqhci_set_tran_desc(*desc, addr, len, end, dma64);
> +		return;
> +	}
> +
> +	offset = addr & (SZ_128M - 1);
> +	tmplen = SZ_128M - offset;
> +	cqhci_set_tran_desc(*desc, addr, tmplen, false, dma64);
> +
> +	addr += tmplen;
> +	len -= tmplen;
> +	*desc += cq_host->trans_desc_len;
> +	cqhci_set_tran_desc(*desc, addr, len, end, dma64);
> +}
> +
> +static void dwcmshc_cqhci_dumpregs(struct mmc_host *mmc)
> +{
> +	sdhci_dumpregs(mmc_priv(mmc));
> +}
> +
>  static void dwcmshc_rk3568_set_clock(struct sdhci_host *host, unsigned int clock)
>  {
>  	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> @@ -345,6 +440,7 @@ static const struct sdhci_ops sdhci_dwcmshc_ops = {
>  	.get_max_clock		= dwcmshc_get_max_clock,
>  	.reset			= sdhci_reset,
>  	.adma_write_desc	= dwcmshc_adma_write_desc,
> +	.irq			= dwcmshc_cqe_irq_handler,
>  };
>  
>  static const struct sdhci_ops sdhci_dwcmshc_rk35xx_ops = {
> @@ -379,6 +475,71 @@ static const struct sdhci_pltfm_data sdhci_dwcmshc_rk35xx_pdata = {
>  		   SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
>  };
>  
> +static const struct cqhci_host_ops dwcmshc_cqhci_ops = {
> +	.enable		= dwcmshc_sdhci_cqe_enable,
> +	.disable	= sdhci_cqe_disable,
> +	.dumpregs	= dwcmshc_cqhci_dumpregs,
> +	.set_tran_desc	= dwcmshc_set_tran_desc,
> +};
> +
> +static void dwcmshc_cqhci_init(struct sdhci_host *host, struct platform_device *pdev)
> +{
> +	struct cqhci_host *cq_host;
> +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> +	struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
> +	bool dma64 = false;
> +	u16 clk;
> +	int err;
> +
> +	host->mmc->caps2 |= MMC_CAP2_CQE | MMC_CAP2_CQE_DCMD;
> +	cq_host = devm_kzalloc(&pdev->dev, sizeof(*cq_host), GFP_KERNEL);
> +	if (!cq_host) {
> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: not enough memory\n");
> +		return;
> +	}
> +
> +	/*
> +	 * For dwcmshc host controller we have to enable internal clock
> +	 * before access to some registers from Vendor Specific Aria 2.

Aria -> Area

> +	 */
> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
> +	clk |= SDHCI_CLOCK_INT_EN;
> +	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
> +	if (!(clk & SDHCI_CLOCK_INT_EN)) {
> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: internal clock enable error\n");
> +		goto free_cq_host;
> +	}
> +
> +	cq_host->mmio = host->ioaddr + priv->vendor_specific_area2;
> +	cq_host->ops = &dwcmshc_cqhci_ops;
> +
> +	/* Enable using of 128-bit task descriptors */
> +	dma64 = host->flags & SDHCI_USE_64_BIT_DMA;
> +	if (dma64) {
> +		dev_dbg(mmc_dev(host->mmc), "128-bit task descriptors\n");
> +		cq_host->caps |= CQHCI_TASK_DESC_SZ_128;
> +	}
> +	err = cqhci_init(cq_host, host->mmc, dma64);
> +	if (err) {
> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: error %d\n", err);
> +		goto int_clock_disable;
> +	}
> +
> +	dev_dbg(mmc_dev(host->mmc), "CQE init done\n");
> +
> +	return;
> +
> +int_clock_disable:
> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
> +	clk &= ~SDHCI_CLOCK_INT_EN;
> +	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
> +
> +free_cq_host:
> +	devm_kfree(&pdev->dev, cq_host);
> +}
> +
> +
>  static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct dwcmshc_priv *dwc_priv)
>  {
>  	int err;
> @@ -471,7 +632,7 @@ static int dwcmshc_probe(struct platform_device *pdev)
>  	struct rk35xx_priv *rk_priv = NULL;
>  	const struct sdhci_pltfm_data *pltfm_data;
>  	int err;
> -	u32 extra;
> +	u32 extra, caps;
>  
>  	pltfm_data = device_get_match_data(&pdev->dev);
>  	if (!pltfm_data) {
> @@ -519,9 +680,12 @@ static int dwcmshc_probe(struct platform_device *pdev)
>  
>  	priv->vendor_specific_area1 =
>  		sdhci_readl(host, DWCMSHC_P_VENDOR_AREA1) & DWCMSHC_AREA1_MASK;
> +	priv->vendor_specific_area2 =
> +		sdhci_readw(host, DWCMSHC_P_VENDOR_AREA2);

Is this OK for all IPs? ie. do they all have DWCMSHC_P_VENDOR_AREA2 register?

>  
>  	host->mmc_host_ops.request = dwcmshc_request;
>  	host->mmc_host_ops.hs400_enhanced_strobe = dwcmshc_hs400_enhanced_strobe;
> +	host->mmc_host_ops.execute_tuning = dwcmshc_execute_tuning;
>  
>  	if (pltfm_data == &sdhci_dwcmshc_rk35xx_pdata) {
>  		rk_priv = devm_kzalloc(&pdev->dev, sizeof(struct rk35xx_priv), GFP_KERNEL);
> @@ -547,6 +711,10 @@ static int dwcmshc_probe(struct platform_device *pdev)
>  		sdhci_enable_v4_mode(host);
>  #endif
>  
> +	caps = sdhci_readl(host, SDHCI_CAPABILITIES);
> +	if (caps & SDHCI_CAN_64BIT_V4)
> +		sdhci_enable_v4_mode(host);
> +
>  	host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
>  
>  	pm_runtime_get_noresume(dev);
> @@ -557,6 +725,15 @@ static int dwcmshc_probe(struct platform_device *pdev)
>  	if (err)
>  		goto err_rpm;
>  
> +	/* Setup Command Queue Engine if enabled */
> +	if (device_property_read_bool(&pdev->dev, "supports-cqe")) {
> +		if (caps & SDHCI_CAN_64BIT_V4)
> +			dwcmshc_cqhci_init(host, pdev);
> +		else
> +			dev_warn(dev, "Cannot enable CQE without V4 mode support\n");
> +	}
> +
> +

Double blank line.

>  	if (rk_priv)
>  		dwcmshc_rk35xx_postinit(host, priv);
>  

Re: [PATCH v5 2/2] mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

[Sergey Khimich]

Hello Adrian!

Sorry for the late reply. Thanks for the comments - I`ll fix the next 
patch set version.

On 03.01.2024 21:52, Adrian Hunter wrote:
> On 31/12/23 16:46, Sergey Khimich wrote:
>> From: Sergey Khimich <serghox@gmail.com>
>>
>> For enabling CQE support just set 'supports-cqe' in your DevTree file
>> for appropriate mmc node.
>>
>> Signed-off-by: Sergey Khimich <serghox@gmail.com>
>> ---
>>   drivers/mmc/host/Kconfig            |   1 +
>>   drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
>>   2 files changed, 180 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
>> index 58bd5fe4cd25..f7594705b013 100644
>> --- a/drivers/mmc/host/Kconfig
>> +++ b/drivers/mmc/host/Kconfig
>> @@ -233,6 +233,7 @@ config MMC_SDHCI_OF_DWCMSHC
>>   	depends on MMC_SDHCI_PLTFM
>>   	depends on OF
>>   	depends on COMMON_CLK
>> +	select MMC_CQHCI
>>   	help
>>   	  This selects Synopsys DesignWare Cores Mobile Storage Controller
>>   	  support.
>> diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
>> index 3a3bae6948a8..0ba1df4bcf36 100644
>> --- a/drivers/mmc/host/sdhci-of-dwcmshc.c
>> +++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
>> @@ -20,6 +20,7 @@
>>   #include <linux/sizes.h>
>>   
>>   #include "sdhci-pltfm.h"
>> +#include "cqhci.h"
>>   
>>   #define SDHCI_DWCMSHC_ARG2_STUFF	GENMASK(31, 16)
>>   
>> @@ -36,6 +37,9 @@
>>   #define DWCMSHC_ENHANCED_STROBE		BIT(8)
>>   #define DWCMSHC_EMMC_ATCTRL		0x40
>>   
>> +/* DWC IP vendor area 2 pointer */
>> +#define DWCMSHC_P_VENDOR_AREA2		0xea
>> +
>>   /* Rockchip specific Registers */
>>   #define DWCMSHC_EMMC_DLL_CTRL		0x800
>>   #define DWCMSHC_EMMC_DLL_RXCLK		0x804
>> @@ -75,6 +79,11 @@
>>   #define BOUNDARY_OK(addr, len) \
>>   	((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
>>   
>> +#define DWCMSHC_SDHCI_CQE_TRNS_MODE	(SDHCI_TRNS_MULTI | \
>> +					 SDHCI_TRNS_BLK_CNT_EN | \
>> +					 SDHCI_TRNS_DMA)
>> +
>> +
>>   enum dwcmshc_rk_type {
>>   	DWCMSHC_RK3568,
>>   	DWCMSHC_RK3588,
>> @@ -90,7 +99,9 @@ struct rk35xx_priv {
>>   
>>   struct dwcmshc_priv {
>>   	struct clk	*bus_clk;
>> -	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA reg */
>> +	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA1 reg */
>> +	int vendor_specific_area2; /* P_VENDOR_SPECIFIC_AREA2 reg */
>> +
>>   	void *priv; /* pointer to SoC private stuff */
>>   };
>>   
>> @@ -210,6 +221,90 @@ static void dwcmshc_hs400_enhanced_strobe(struct mmc_host *mmc,
>>   	sdhci_writel(host, vendor, reg);
>>   }
>>   
>> +static int dwcmshc_execute_tuning(struct mmc_host *mmc, u32 opcode)
>> +{
>> +	int err = sdhci_execute_tuning(mmc, opcode);
>> +	struct sdhci_host *host = mmc_priv(mmc);
>> +
>> +	if (err)
>> +		return err;
>> +
>> +	/*
>> +	 * Tuning can leave the IP in an active state (Buffer Read Enable bit
>> +	 * set) which prevents the entry to low power states (i.e. S0i3). Data
>> +	 * reset will clear it.
>> +	 */
>> +	sdhci_reset(host, SDHCI_RESET_DATA);
>> +
>> +	return 0;
>> +}
>> +
>> +static u32 dwcmshc_cqe_irq_handler(struct sdhci_host *host, u32 intmask)
>> +{
>> +	int cmd_error = 0;
>> +	int data_error = 0;
>> +
>> +	if (!sdhci_cqe_irq(host, intmask, &cmd_error, &data_error))
>> +		return intmask;
>> +
>> +	cqhci_irq(host->mmc, intmask, cmd_error, data_error);
>> +
>> +	return 0;
>> +}
>> +
>> +static void dwcmshc_sdhci_cqe_enable(struct mmc_host *mmc)
>> +{
>> +	struct sdhci_host *host = mmc_priv(mmc);
>> +	u8 ctrl;
>> +
>> +	sdhci_writew(host, DWCMSHC_SDHCI_CQE_TRNS_MODE, SDHCI_TRANSFER_MODE);
>> +
>> +	sdhci_cqe_enable(mmc);
>> +
>> +	/*
>> +	 * The "DesignWare Cores Mobile Storage Host Controller
>> +	 * DWC_mshc / DWC_mshc_lite Databook" says:
>> +	 * when Host Version 4 Enable" is 1 in Host Control 2 register,
>> +	 * SDHCI_CTRL_ADMA32 bit means ADMA2 is selected.
>> +	 * Selection of 32-bit/64-bit System Addressing:
>> +	 * either 32-bit or 64-bit system addressing is selected by
>> +	 * 64-bit Addressing bit in Host Control 2 register.
>> +	 *
>> +	 * On the other hand the "DesignWare Cores Mobile Storage Host
>> +	 * Controller DWC_mshc / DWC_mshc_lite User Guide" says, that we have to
>> +	 * set DMA_SEL to ADMA2 _only_ mode in the Host Control 2 register.
>> +	 */
>> +	ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
>> +	ctrl &= ~SDHCI_CTRL_DMA_MASK;
>> +	ctrl |= SDHCI_CTRL_ADMA32;
>> +	sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
>> +}
>> +
>> +static void dwcmshc_set_tran_desc(struct cqhci_host *cq_host, u8 **desc,
>> +				  dma_addr_t addr, int len, bool end, bool dma64)
>> +{
>> +	int tmplen, offset;
>> +
>> +	if (likely(!len || BOUNDARY_OK(addr, len))) {
>> +		cqhci_set_tran_desc(*desc, addr, len, end, dma64);
>> +		return;
>> +	}
>> +
>> +	offset = addr & (SZ_128M - 1);
>> +	tmplen = SZ_128M - offset;
>> +	cqhci_set_tran_desc(*desc, addr, tmplen, false, dma64);
>> +
>> +	addr += tmplen;
>> +	len -= tmplen;
>> +	*desc += cq_host->trans_desc_len;
>> +	cqhci_set_tran_desc(*desc, addr, len, end, dma64);
>> +}
>> +
>> +static void dwcmshc_cqhci_dumpregs(struct mmc_host *mmc)
>> +{
>> +	sdhci_dumpregs(mmc_priv(mmc));
>> +}
>> +
>>   static void dwcmshc_rk3568_set_clock(struct sdhci_host *host, unsigned int clock)
>>   {
>>   	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> @@ -345,6 +440,7 @@ static const struct sdhci_ops sdhci_dwcmshc_ops = {
>>   	.get_max_clock		= dwcmshc_get_max_clock,
>>   	.reset			= sdhci_reset,
>>   	.adma_write_desc	= dwcmshc_adma_write_desc,
>> +	.irq			= dwcmshc_cqe_irq_handler,
>>   };
>>   
>>   static const struct sdhci_ops sdhci_dwcmshc_rk35xx_ops = {
>> @@ -379,6 +475,71 @@ static const struct sdhci_pltfm_data sdhci_dwcmshc_rk35xx_pdata = {
>>   		   SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
>>   };
>>   
>> +static const struct cqhci_host_ops dwcmshc_cqhci_ops = {
>> +	.enable		= dwcmshc_sdhci_cqe_enable,
>> +	.disable	= sdhci_cqe_disable,
>> +	.dumpregs	= dwcmshc_cqhci_dumpregs,
>> +	.set_tran_desc	= dwcmshc_set_tran_desc,
>> +};
>> +
>> +static void dwcmshc_cqhci_init(struct sdhci_host *host, struct platform_device *pdev)
>> +{
>> +	struct cqhci_host *cq_host;
>> +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> +	struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
>> +	bool dma64 = false;
>> +	u16 clk;
>> +	int err;
>> +
>> +	host->mmc->caps2 |= MMC_CAP2_CQE | MMC_CAP2_CQE_DCMD;
>> +	cq_host = devm_kzalloc(&pdev->dev, sizeof(*cq_host), GFP_KERNEL);
>> +	if (!cq_host) {
>> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: not enough memory\n");
>> +		return;
>> +	}
>> +
>> +	/*
>> +	 * For dwcmshc host controller we have to enable internal clock
>> +	 * before access to some registers from Vendor Specific Aria 2.
> Aria -> Area
>
>> +	 */
>> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>> +	clk |= SDHCI_CLOCK_INT_EN;
>> +	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
>> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>> +	if (!(clk & SDHCI_CLOCK_INT_EN)) {
>> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: internal clock enable error\n");
>> +		goto free_cq_host;
>> +	}
>> +
>> +	cq_host->mmio = host->ioaddr + priv->vendor_specific_area2;
>> +	cq_host->ops = &dwcmshc_cqhci_ops;
>> +
>> +	/* Enable using of 128-bit task descriptors */
>> +	dma64 = host->flags & SDHCI_USE_64_BIT_DMA;
>> +	if (dma64) {
>> +		dev_dbg(mmc_dev(host->mmc), "128-bit task descriptors\n");
>> +		cq_host->caps |= CQHCI_TASK_DESC_SZ_128;
>> +	}
>> +	err = cqhci_init(cq_host, host->mmc, dma64);
>> +	if (err) {
>> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: error %d\n", err);
>> +		goto int_clock_disable;
>> +	}
>> +
>> +	dev_dbg(mmc_dev(host->mmc), "CQE init done\n");
>> +
>> +	return;
>> +
>> +int_clock_disable:
>> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>> +	clk &= ~SDHCI_CLOCK_INT_EN;
>> +	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
>> +
>> +free_cq_host:
>> +	devm_kfree(&pdev->dev, cq_host);
>> +}
>> +
>> +
>>   static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct dwcmshc_priv *dwc_priv)
>>   {
>>   	int err;
>> @@ -471,7 +632,7 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>   	struct rk35xx_priv *rk_priv = NULL;
>>   	const struct sdhci_pltfm_data *pltfm_data;
>>   	int err;
>> -	u32 extra;
>> +	u32 extra, caps;
>>   
>>   	pltfm_data = device_get_match_data(&pdev->dev);
>>   	if (!pltfm_data) {
>> @@ -519,9 +680,12 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>   
>>   	priv->vendor_specific_area1 =
>>   		sdhci_readl(host, DWCMSHC_P_VENDOR_AREA1) & DWCMSHC_AREA1_MASK;
>> +	priv->vendor_specific_area2 =
>> +		sdhci_readw(host, DWCMSHC_P_VENDOR_AREA2);
> Is this OK for all IPs? ie. do they all have DWCMSHC_P_VENDOR_AREA2 register?
Good question. Actually I can't guarantee that all dwcmshc-like IPs have 
DWCMSHC_P_VENDOR_AREA2 register,
because I`ve tested my patch only on my pre-selicon FPGA prototype.
But on the other hand according to the documentation I have, CQE-related 
registers are placed within this
vendor2 block register. And thanks to Shawn Lin we know, Rockchip SoCs 
with this IP also have this reg.
As I understand CQE for dwcmshc-like IPs will not work without 
DWCMSHC_P_VENDOR_AREA2.
>
>>   
>>   	host->mmc_host_ops.request = dwcmshc_request;
>>   	host->mmc_host_ops.hs400_enhanced_strobe = dwcmshc_hs400_enhanced_strobe;
>> +	host->mmc_host_ops.execute_tuning = dwcmshc_execute_tuning;
>>   
>>   	if (pltfm_data == &sdhci_dwcmshc_rk35xx_pdata) {
>>   		rk_priv = devm_kzalloc(&pdev->dev, sizeof(struct rk35xx_priv), GFP_KERNEL);
>> @@ -547,6 +711,10 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>   		sdhci_enable_v4_mode(host);
>>   #endif
>>   
>> +	caps = sdhci_readl(host, SDHCI_CAPABILITIES);
>> +	if (caps & SDHCI_CAN_64BIT_V4)
>> +		sdhci_enable_v4_mode(host);
>> +
>>   	host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
>>   
>>   	pm_runtime_get_noresume(dev);
>> @@ -557,6 +725,15 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>   	if (err)
>>   		goto err_rpm;
>>   
>> +	/* Setup Command Queue Engine if enabled */
>> +	if (device_property_read_bool(&pdev->dev, "supports-cqe")) {
>> +		if (caps & SDHCI_CAN_64BIT_V4)
>> +			dwcmshc_cqhci_init(host, pdev);
>> +		else
>> +			dev_warn(dev, "Cannot enable CQE without V4 mode support\n");
>> +	}
>> +
>> +
> Double blank line.
>
>>   	if (rk_priv)
>>   		dwcmshc_rk35xx_postinit(host, priv);
>>   

Re: [PATCH v5 2/2] mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

[Adrian Hunter]

On 4/03/24 16:30, Sergey Khimich wrote:
> Hello Adrian!
> 
> Sorry for the late reply. Thanks for the comments - I`ll fix the next patch set version.
> 
> On 03.01.2024 21:52, Adrian Hunter wrote:
>> On 31/12/23 16:46, Sergey Khimich wrote:
>>> From: Sergey Khimich <serghox@gmail.com>
>>>
>>> For enabling CQE support just set 'supports-cqe' in your DevTree file
>>> for appropriate mmc node.
>>>
>>> Signed-off-by: Sergey Khimich <serghox@gmail.com>
>>> ---
>>>   drivers/mmc/host/Kconfig            |   1 +
>>>   drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
>>>   2 files changed, 180 insertions(+), 2 deletions(-)
>>>
>>> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
>>> index 58bd5fe4cd25..f7594705b013 100644
>>> --- a/drivers/mmc/host/Kconfig
>>> +++ b/drivers/mmc/host/Kconfig
>>> @@ -233,6 +233,7 @@ config MMC_SDHCI_OF_DWCMSHC
>>>       depends on MMC_SDHCI_PLTFM
>>>       depends on OF
>>>       depends on COMMON_CLK
>>> +    select MMC_CQHCI
>>>       help
>>>         This selects Synopsys DesignWare Cores Mobile Storage Controller
>>>         support.
>>> diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
>>> index 3a3bae6948a8..0ba1df4bcf36 100644
>>> --- a/drivers/mmc/host/sdhci-of-dwcmshc.c
>>> +++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
>>> @@ -20,6 +20,7 @@
>>>   #include <linux/sizes.h>
>>>     #include "sdhci-pltfm.h"
>>> +#include "cqhci.h"
>>>     #define SDHCI_DWCMSHC_ARG2_STUFF    GENMASK(31, 16)
>>>   @@ -36,6 +37,9 @@
>>>   #define DWCMSHC_ENHANCED_STROBE        BIT(8)
>>>   #define DWCMSHC_EMMC_ATCTRL        0x40
>>>   +/* DWC IP vendor area 2 pointer */
>>> +#define DWCMSHC_P_VENDOR_AREA2        0xea
>>> +
>>>   /* Rockchip specific Registers */
>>>   #define DWCMSHC_EMMC_DLL_CTRL        0x800
>>>   #define DWCMSHC_EMMC_DLL_RXCLK        0x804
>>> @@ -75,6 +79,11 @@
>>>   #define BOUNDARY_OK(addr, len) \
>>>       ((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
>>>   +#define DWCMSHC_SDHCI_CQE_TRNS_MODE    (SDHCI_TRNS_MULTI | \
>>> +                     SDHCI_TRNS_BLK_CNT_EN | \
>>> +                     SDHCI_TRNS_DMA)
>>> +
>>> +
>>>   enum dwcmshc_rk_type {
>>>       DWCMSHC_RK3568,
>>>       DWCMSHC_RK3588,
>>> @@ -90,7 +99,9 @@ struct rk35xx_priv {
>>>     struct dwcmshc_priv {
>>>       struct clk    *bus_clk;
>>> -    int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA reg */
>>> +    int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA1 reg */
>>> +    int vendor_specific_area2; /* P_VENDOR_SPECIFIC_AREA2 reg */
>>> +
>>>       void *priv; /* pointer to SoC private stuff */
>>>   };
>>>   @@ -210,6 +221,90 @@ static void dwcmshc_hs400_enhanced_strobe(struct mmc_host *mmc,
>>>       sdhci_writel(host, vendor, reg);
>>>   }
>>>   +static int dwcmshc_execute_tuning(struct mmc_host *mmc, u32 opcode)
>>> +{
>>> +    int err = sdhci_execute_tuning(mmc, opcode);
>>> +    struct sdhci_host *host = mmc_priv(mmc);
>>> +
>>> +    if (err)
>>> +        return err;
>>> +
>>> +    /*
>>> +     * Tuning can leave the IP in an active state (Buffer Read Enable bit
>>> +     * set) which prevents the entry to low power states (i.e. S0i3). Data
>>> +     * reset will clear it.
>>> +     */
>>> +    sdhci_reset(host, SDHCI_RESET_DATA);
>>> +
>>> +    return 0;
>>> +}
>>> +
>>> +static u32 dwcmshc_cqe_irq_handler(struct sdhci_host *host, u32 intmask)
>>> +{
>>> +    int cmd_error = 0;
>>> +    int data_error = 0;
>>> +
>>> +    if (!sdhci_cqe_irq(host, intmask, &cmd_error, &data_error))
>>> +        return intmask;
>>> +
>>> +    cqhci_irq(host->mmc, intmask, cmd_error, data_error);
>>> +
>>> +    return 0;
>>> +}
>>> +
>>> +static void dwcmshc_sdhci_cqe_enable(struct mmc_host *mmc)
>>> +{
>>> +    struct sdhci_host *host = mmc_priv(mmc);
>>> +    u8 ctrl;
>>> +
>>> +    sdhci_writew(host, DWCMSHC_SDHCI_CQE_TRNS_MODE, SDHCI_TRANSFER_MODE);
>>> +
>>> +    sdhci_cqe_enable(mmc);
>>> +
>>> +    /*
>>> +     * The "DesignWare Cores Mobile Storage Host Controller
>>> +     * DWC_mshc / DWC_mshc_lite Databook" says:
>>> +     * when Host Version 4 Enable" is 1 in Host Control 2 register,
>>> +     * SDHCI_CTRL_ADMA32 bit means ADMA2 is selected.
>>> +     * Selection of 32-bit/64-bit System Addressing:
>>> +     * either 32-bit or 64-bit system addressing is selected by
>>> +     * 64-bit Addressing bit in Host Control 2 register.
>>> +     *
>>> +     * On the other hand the "DesignWare Cores Mobile Storage Host
>>> +     * Controller DWC_mshc / DWC_mshc_lite User Guide" says, that we have to
>>> +     * set DMA_SEL to ADMA2 _only_ mode in the Host Control 2 register.
>>> +     */
>>> +    ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
>>> +    ctrl &= ~SDHCI_CTRL_DMA_MASK;
>>> +    ctrl |= SDHCI_CTRL_ADMA32;
>>> +    sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
>>> +}
>>> +
>>> +static void dwcmshc_set_tran_desc(struct cqhci_host *cq_host, u8 **desc,
>>> +                  dma_addr_t addr, int len, bool end, bool dma64)
>>> +{
>>> +    int tmplen, offset;
>>> +
>>> +    if (likely(!len || BOUNDARY_OK(addr, len))) {
>>> +        cqhci_set_tran_desc(*desc, addr, len, end, dma64);
>>> +        return;
>>> +    }
>>> +
>>> +    offset = addr & (SZ_128M - 1);
>>> +    tmplen = SZ_128M - offset;
>>> +    cqhci_set_tran_desc(*desc, addr, tmplen, false, dma64);
>>> +
>>> +    addr += tmplen;
>>> +    len -= tmplen;
>>> +    *desc += cq_host->trans_desc_len;
>>> +    cqhci_set_tran_desc(*desc, addr, len, end, dma64);
>>> +}
>>> +
>>> +static void dwcmshc_cqhci_dumpregs(struct mmc_host *mmc)
>>> +{
>>> +    sdhci_dumpregs(mmc_priv(mmc));
>>> +}
>>> +
>>>   static void dwcmshc_rk3568_set_clock(struct sdhci_host *host, unsigned int clock)
>>>   {
>>>       struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>>> @@ -345,6 +440,7 @@ static const struct sdhci_ops sdhci_dwcmshc_ops = {
>>>       .get_max_clock        = dwcmshc_get_max_clock,
>>>       .reset            = sdhci_reset,
>>>       .adma_write_desc    = dwcmshc_adma_write_desc,
>>> +    .irq            = dwcmshc_cqe_irq_handler,
>>>   };
>>>     static const struct sdhci_ops sdhci_dwcmshc_rk35xx_ops = {
>>> @@ -379,6 +475,71 @@ static const struct sdhci_pltfm_data sdhci_dwcmshc_rk35xx_pdata = {
>>>              SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
>>>   };
>>>   +static const struct cqhci_host_ops dwcmshc_cqhci_ops = {
>>> +    .enable        = dwcmshc_sdhci_cqe_enable,
>>> +    .disable    = sdhci_cqe_disable,
>>> +    .dumpregs    = dwcmshc_cqhci_dumpregs,
>>> +    .set_tran_desc    = dwcmshc_set_tran_desc,
>>> +};
>>> +
>>> +static void dwcmshc_cqhci_init(struct sdhci_host *host, struct platform_device *pdev)
>>> +{
>>> +    struct cqhci_host *cq_host;
>>> +    struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>>> +    struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
>>> +    bool dma64 = false;
>>> +    u16 clk;
>>> +    int err;
>>> +
>>> +    host->mmc->caps2 |= MMC_CAP2_CQE | MMC_CAP2_CQE_DCMD;
>>> +    cq_host = devm_kzalloc(&pdev->dev, sizeof(*cq_host), GFP_KERNEL);
>>> +    if (!cq_host) {
>>> +        dev_err(mmc_dev(host->mmc), "Unable to setup CQE: not enough memory\n");
>>> +        return;
>>> +    }
>>> +
>>> +    /*
>>> +     * For dwcmshc host controller we have to enable internal clock
>>> +     * before access to some registers from Vendor Specific Aria 2.
>> Aria -> Area
>>
>>> +     */
>>> +    clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>>> +    clk |= SDHCI_CLOCK_INT_EN;
>>> +    sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
>>> +    clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>>> +    if (!(clk & SDHCI_CLOCK_INT_EN)) {
>>> +        dev_err(mmc_dev(host->mmc), "Unable to setup CQE: internal clock enable error\n");
>>> +        goto free_cq_host;
>>> +    }
>>> +
>>> +    cq_host->mmio = host->ioaddr + priv->vendor_specific_area2;
>>> +    cq_host->ops = &dwcmshc_cqhci_ops;
>>> +
>>> +    /* Enable using of 128-bit task descriptors */
>>> +    dma64 = host->flags & SDHCI_USE_64_BIT_DMA;
>>> +    if (dma64) {
>>> +        dev_dbg(mmc_dev(host->mmc), "128-bit task descriptors\n");
>>> +        cq_host->caps |= CQHCI_TASK_DESC_SZ_128;
>>> +    }
>>> +    err = cqhci_init(cq_host, host->mmc, dma64);
>>> +    if (err) {
>>> +        dev_err(mmc_dev(host->mmc), "Unable to setup CQE: error %d\n", err);
>>> +        goto int_clock_disable;
>>> +    }
>>> +
>>> +    dev_dbg(mmc_dev(host->mmc), "CQE init done\n");
>>> +
>>> +    return;
>>> +
>>> +int_clock_disable:
>>> +    clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>>> +    clk &= ~SDHCI_CLOCK_INT_EN;
>>> +    sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
>>> +
>>> +free_cq_host:
>>> +    devm_kfree(&pdev->dev, cq_host);
>>> +}
>>> +
>>> +
>>>   static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct dwcmshc_priv *dwc_priv)
>>>   {
>>>       int err;
>>> @@ -471,7 +632,7 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>>       struct rk35xx_priv *rk_priv = NULL;
>>>       const struct sdhci_pltfm_data *pltfm_data;
>>>       int err;
>>> -    u32 extra;
>>> +    u32 extra, caps;
>>>         pltfm_data = device_get_match_data(&pdev->dev);
>>>       if (!pltfm_data) {
>>> @@ -519,9 +680,12 @@ static int dwcmshc_probe(struct platform_device *pdev)
>>>         priv->vendor_specific_area1 =
>>>           sdhci_readl(host, DWCMSHC_P_VENDOR_AREA1) & DWCMSHC_AREA1_MASK;
>>> +    priv->vendor_specific_area2 =
>>> +        sdhci_readw(host, DWCMSHC_P_VENDOR_AREA2);
>> Is this OK for all IPs? ie. do they all have DWCMSHC_P_VENDOR_AREA2 register?
> Good question. Actually I can't guarantee that all dwcmshc-like IPs have DWCMSHC_P_VENDOR_AREA2 register,
> because I`ve tested my patch only on my pre-selicon FPGA prototype.
> But on the other hand according to the documentation I have, CQE-related registers are placed within this
> vendor2 block register. And thanks to Shawn Lin we know, Rockchip SoCs with this IP also have this reg.
> As I understand CQE for dwcmshc-like IPs will not work without DWCMSHC_P_VENDOR_AREA2.

Maybe it should be read only in that case? For example, after:

	if (device_property_read_bool(&pdev->dev, "supports-cqe")) {

Re: [PATCH v5 2/2] mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

[Shawn Lin]

Hi Sergey

On 2023/12/31 22:46, Sergey Khimich wrote:
> From: Sergey Khimich <serghox@gmail.com>
> 
> For enabling CQE support just set 'supports-cqe' in your DevTree file
> for appropriate mmc node.
> 

I have tested this patchset and it seems work fine on Rockchip SoCs when 
I removed sdhci_enable_v4_mode as we don't have SDHCI_CAN_64BIT_V4 
support. Why do we force to enable v4?

And I found it breaks s2r, since the patch didn't call cqhci_suspend
and cqhci_resume.

> Signed-off-by: Sergey Khimich <serghox@gmail.com>
> ---
>   drivers/mmc/host/Kconfig            |   1 +
>   drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
>   2 files changed, 180 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
> index 58bd5fe4cd25..f7594705b013 100644
> --- a/drivers/mmc/host/Kconfig
> +++ b/drivers/mmc/host/Kconfig
> @@ -233,6 +233,7 @@ config MMC_SDHCI_OF_DWCMSHC
>   	depends on MMC_SDHCI_PLTFM
>   	depends on OF
>   	depends on COMMON_CLK
> +	select MMC_CQHCI
>   	help
>   	  This selects Synopsys DesignWare Cores Mobile Storage Controller
>   	  support.
> diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c b/drivers/mmc/host/sdhci-of-dwcmshc.c
> index 3a3bae6948a8..0ba1df4bcf36 100644
> --- a/drivers/mmc/host/sdhci-of-dwcmshc.c
> +++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
> @@ -20,6 +20,7 @@
>   #include <linux/sizes.h>
>   
>   #include "sdhci-pltfm.h"
> +#include "cqhci.h"
>   
>   #define SDHCI_DWCMSHC_ARG2_STUFF	GENMASK(31, 16)
>   
> @@ -36,6 +37,9 @@
>   #define DWCMSHC_ENHANCED_STROBE		BIT(8)
>   #define DWCMSHC_EMMC_ATCTRL		0x40
>   
> +/* DWC IP vendor area 2 pointer */
> +#define DWCMSHC_P_VENDOR_AREA2		0xea
> +
>   /* Rockchip specific Registers */
>   #define DWCMSHC_EMMC_DLL_CTRL		0x800
>   #define DWCMSHC_EMMC_DLL_RXCLK		0x804
> @@ -75,6 +79,11 @@
>   #define BOUNDARY_OK(addr, len) \
>   	((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
>   
> +#define DWCMSHC_SDHCI_CQE_TRNS_MODE	(SDHCI_TRNS_MULTI | \
> +					 SDHCI_TRNS_BLK_CNT_EN | \
> +					 SDHCI_TRNS_DMA)
> +
> +
>   enum dwcmshc_rk_type {
>   	DWCMSHC_RK3568,
>   	DWCMSHC_RK3588,
> @@ -90,7 +99,9 @@ struct rk35xx_priv {
>   
>   struct dwcmshc_priv {
>   	struct clk	*bus_clk;
> -	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA reg */
> +	int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA1 reg */
> +	int vendor_specific_area2; /* P_VENDOR_SPECIFIC_AREA2 reg */
> +
>   	void *priv; /* pointer to SoC private stuff */
>   };
>   
> @@ -210,6 +221,90 @@ static void dwcmshc_hs400_enhanced_strobe(struct mmc_host *mmc,
>   	sdhci_writel(host, vendor, reg);
>   }
>   
> +static int dwcmshc_execute_tuning(struct mmc_host *mmc, u32 opcode)
> +{
> +	int err = sdhci_execute_tuning(mmc, opcode);
> +	struct sdhci_host *host = mmc_priv(mmc);
> +
> +	if (err)
> +		return err;
> +
> +	/*
> +	 * Tuning can leave the IP in an active state (Buffer Read Enable bit
> +	 * set) which prevents the entry to low power states (i.e. S0i3). Data
> +	 * reset will clear it.
> +	 */
> +	sdhci_reset(host, SDHCI_RESET_DATA);
> +
> +	return 0;
> +}
> +
> +static u32 dwcmshc_cqe_irq_handler(struct sdhci_host *host, u32 intmask)
> +{
> +	int cmd_error = 0;
> +	int data_error = 0;
> +
> +	if (!sdhci_cqe_irq(host, intmask, &cmd_error, &data_error))
> +		return intmask;
> +
> +	cqhci_irq(host->mmc, intmask, cmd_error, data_error);
> +
> +	return 0;
> +}
> +
> +static void dwcmshc_sdhci_cqe_enable(struct mmc_host *mmc)
> +{
> +	struct sdhci_host *host = mmc_priv(mmc);
> +	u8 ctrl;
> +
> +	sdhci_writew(host, DWCMSHC_SDHCI_CQE_TRNS_MODE, SDHCI_TRANSFER_MODE);
> +
> +	sdhci_cqe_enable(mmc);
> +
> +	/*
> +	 * The "DesignWare Cores Mobile Storage Host Controller
> +	 * DWC_mshc / DWC_mshc_lite Databook" says:
> +	 * when Host Version 4 Enable" is 1 in Host Control 2 register,
> +	 * SDHCI_CTRL_ADMA32 bit means ADMA2 is selected.
> +	 * Selection of 32-bit/64-bit System Addressing:
> +	 * either 32-bit or 64-bit system addressing is selected by
> +	 * 64-bit Addressing bit in Host Control 2 register.
> +	 *
> +	 * On the other hand the "DesignWare Cores Mobile Storage Host
> +	 * Controller DWC_mshc / DWC_mshc_lite User Guide" says, that we have to
> +	 * set DMA_SEL to ADMA2 _only_ mode in the Host Control 2 register.
> +	 */
> +	ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
> +	ctrl &= ~SDHCI_CTRL_DMA_MASK;
> +	ctrl |= SDHCI_CTRL_ADMA32;
> +	sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
> +}
> +
> +static void dwcmshc_set_tran_desc(struct cqhci_host *cq_host, u8 **desc,
> +				  dma_addr_t addr, int len, bool end, bool dma64)
> +{
> +	int tmplen, offset;
> +
> +	if (likely(!len || BOUNDARY_OK(addr, len))) {
> +		cqhci_set_tran_desc(*desc, addr, len, end, dma64);
> +		return;
> +	}
> +
> +	offset = addr & (SZ_128M - 1);
> +	tmplen = SZ_128M - offset;
> +	cqhci_set_tran_desc(*desc, addr, tmplen, false, dma64);
> +
> +	addr += tmplen;
> +	len -= tmplen;
> +	*desc += cq_host->trans_desc_len;
> +	cqhci_set_tran_desc(*desc, addr, len, end, dma64);
> +}
> +
> +static void dwcmshc_cqhci_dumpregs(struct mmc_host *mmc)
> +{
> +	sdhci_dumpregs(mmc_priv(mmc));
> +}
> +
>   static void dwcmshc_rk3568_set_clock(struct sdhci_host *host, unsigned int clock)
>   {
>   	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> @@ -345,6 +440,7 @@ static const struct sdhci_ops sdhci_dwcmshc_ops = {
>   	.get_max_clock		= dwcmshc_get_max_clock,
>   	.reset			= sdhci_reset,
>   	.adma_write_desc	= dwcmshc_adma_write_desc,
> +	.irq			= dwcmshc_cqe_irq_handler,
>   };
>   
>   static const struct sdhci_ops sdhci_dwcmshc_rk35xx_ops = {
> @@ -379,6 +475,71 @@ static const struct sdhci_pltfm_data sdhci_dwcmshc_rk35xx_pdata = {
>   		   SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
>   };
>   
> +static const struct cqhci_host_ops dwcmshc_cqhci_ops = {
> +	.enable		= dwcmshc_sdhci_cqe_enable,
> +	.disable	= sdhci_cqe_disable,
> +	.dumpregs	= dwcmshc_cqhci_dumpregs,
> +	.set_tran_desc	= dwcmshc_set_tran_desc,
> +};
> +
> +static void dwcmshc_cqhci_init(struct sdhci_host *host, struct platform_device *pdev)
> +{
> +	struct cqhci_host *cq_host;
> +	struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
> +	struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
> +	bool dma64 = false;
> +	u16 clk;
> +	int err;
> +
> +	host->mmc->caps2 |= MMC_CAP2_CQE | MMC_CAP2_CQE_DCMD;
> +	cq_host = devm_kzalloc(&pdev->dev, sizeof(*cq_host), GFP_KERNEL);
> +	if (!cq_host) {
> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: not enough memory\n");
> +		return;
> +	}
> +
> +	/*
> +	 * For dwcmshc host controller we have to enable internal clock
> +	 * before access to some registers from Vendor Specific Aria 2.
> +	 */
> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
> +	clk |= SDHCI_CLOCK_INT_EN;
> +	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
> +	if (!(clk & SDHCI_CLOCK_INT_EN)) {
> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: internal clock enable error\n");
> +		goto free_cq_host;
> +	}
> +
> +	cq_host->mmio = host->ioaddr + priv->vendor_specific_area2;
> +	cq_host->ops = &dwcmshc_cqhci_ops;
> +
> +	/* Enable using of 128-bit task descriptors */
> +	dma64 = host->flags & SDHCI_USE_64_BIT_DMA;
> +	if (dma64) {
> +		dev_dbg(mmc_dev(host->mmc), "128-bit task descriptors\n");
> +		cq_host->caps |= CQHCI_TASK_DESC_SZ_128;
> +	}
> +	err = cqhci_init(cq_host, host->mmc, dma64);
> +	if (err) {
> +		dev_err(mmc_dev(host->mmc), "Unable to setup CQE: error %d\n", err);
> +		goto int_clock_disable;
> +	}
> +
> +	dev_dbg(mmc_dev(host->mmc), "CQE init done\n");
> +
> +	return;
> +
> +int_clock_disable:
> +	clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
> +	clk &= ~SDHCI_CLOCK_INT_EN;
> +	sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
> +
> +free_cq_host:
> +	devm_kfree(&pdev->dev, cq_host);
> +}
> +
> +
>   static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct dwcmshc_priv *dwc_priv)
>   {
>   	int err;
> @@ -471,7 +632,7 @@ static int dwcmshc_probe(struct platform_device *pdev)
>   	struct rk35xx_priv *rk_priv = NULL;
>   	const struct sdhci_pltfm_data *pltfm_data;
>   	int err;
> -	u32 extra;
> +	u32 extra, caps;
>   
>   	pltfm_data = device_get_match_data(&pdev->dev);
>   	if (!pltfm_data) {
> @@ -519,9 +680,12 @@ static int dwcmshc_probe(struct platform_device *pdev)
>   
>   	priv->vendor_specific_area1 =
>   		sdhci_readl(host, DWCMSHC_P_VENDOR_AREA1) & DWCMSHC_AREA1_MASK;
> +	priv->vendor_specific_area2 =
> +		sdhci_readw(host, DWCMSHC_P_VENDOR_AREA2);
>   
>   	host->mmc_host_ops.request = dwcmshc_request;
>   	host->mmc_host_ops.hs400_enhanced_strobe = dwcmshc_hs400_enhanced_strobe;
> +	host->mmc_host_ops.execute_tuning = dwcmshc_execute_tuning;
>   
>   	if (pltfm_data == &sdhci_dwcmshc_rk35xx_pdata) {
>   		rk_priv = devm_kzalloc(&pdev->dev, sizeof(struct rk35xx_priv), GFP_KERNEL);
> @@ -547,6 +711,10 @@ static int dwcmshc_probe(struct platform_device *pdev)
>   		sdhci_enable_v4_mode(host);
>   #endif
>   
> +	caps = sdhci_readl(host, SDHCI_CAPABILITIES);
> +	if (caps & SDHCI_CAN_64BIT_V4)
> +		sdhci_enable_v4_mode(host);
> +
>   	host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
>   
>   	pm_runtime_get_noresume(dev);
> @@ -557,6 +725,15 @@ static int dwcmshc_probe(struct platform_device *pdev)
>   	if (err)
>   		goto err_rpm;
>   
> +	/* Setup Command Queue Engine if enabled */
> +	if (device_property_read_bool(&pdev->dev, "supports-cqe")) {
> +		if (caps & SDHCI_CAN_64BIT_V4)
> +			dwcmshc_cqhci_init(host, pdev);
> +		else
> +			dev_warn(dev, "Cannot enable CQE without V4 mode support\n");
> +	}
> +
> +
>   	if (rk_priv)
>   		dwcmshc_rk35xx_postinit(host, priv);
>   

Re: [PATCH v5 2/2] mmc: sdhci-of-dwcmshc: Implement SDHCI CQE support

[Sergey Khimich]

Hi Shawn!

Sorry for the late reply. And I really appreciate you've tested my patch 
on Rockchip SoCs! I'll add cqhci_suspend/resume to fix work with s2r.

Speaking about force enabling v4 mode - actually I never managed to run 
CQE for my IP with disabled v4 mode and I thought, that it doesn`t work

without v4. Now when you've tested the case I suggest to remove the 
checking of v4 mode for CQE in general, but I'll add force setting v4

for my platform (when it will be ready) in separate patch.

On 04.01.2024 6:01, Shawn Lin wrote:
> Hi Sergey
>
> On 2023/12/31 22:46, Sergey Khimich wrote:
>> From: Sergey Khimich <serghox@gmail.com>
>>
>> For enabling CQE support just set 'supports-cqe' in your DevTree file
>> for appropriate mmc node.
>>
>
> I have tested this patchset and it seems work fine on Rockchip SoCs 
> when I removed sdhci_enable_v4_mode as we don't have 
> SDHCI_CAN_64BIT_V4 support. Why do we force to enable v4?
>
> And I found it breaks s2r, since the patch didn't call cqhci_suspend
> and cqhci_resume.
>
>> Signed-off-by: Sergey Khimich <serghox@gmail.com>
>> ---
>>   drivers/mmc/host/Kconfig            |   1 +
>>   drivers/mmc/host/sdhci-of-dwcmshc.c | 181 +++++++++++++++++++++++++++-
>>   2 files changed, 180 insertions(+), 2 deletions(-)
>>
>> diff --git a/drivers/mmc/host/Kconfig b/drivers/mmc/host/Kconfig
>> index 58bd5fe4cd25..f7594705b013 100644
>> --- a/drivers/mmc/host/Kconfig
>> +++ b/drivers/mmc/host/Kconfig
>> @@ -233,6 +233,7 @@ config MMC_SDHCI_OF_DWCMSHC
>>       depends on MMC_SDHCI_PLTFM
>>       depends on OF
>>       depends on COMMON_CLK
>> +    select MMC_CQHCI
>>       help
>>         This selects Synopsys DesignWare Cores Mobile Storage Controller
>>         support.
>> diff --git a/drivers/mmc/host/sdhci-of-dwcmshc.c 
>> b/drivers/mmc/host/sdhci-of-dwcmshc.c
>> index 3a3bae6948a8..0ba1df4bcf36 100644
>> --- a/drivers/mmc/host/sdhci-of-dwcmshc.c
>> +++ b/drivers/mmc/host/sdhci-of-dwcmshc.c
>> @@ -20,6 +20,7 @@
>>   #include <linux/sizes.h>
>>     #include "sdhci-pltfm.h"
>> +#include "cqhci.h"
>>     #define SDHCI_DWCMSHC_ARG2_STUFF    GENMASK(31, 16)
>>   @@ -36,6 +37,9 @@
>>   #define DWCMSHC_ENHANCED_STROBE        BIT(8)
>>   #define DWCMSHC_EMMC_ATCTRL        0x40
>>   +/* DWC IP vendor area 2 pointer */
>> +#define DWCMSHC_P_VENDOR_AREA2        0xea
>> +
>>   /* Rockchip specific Registers */
>>   #define DWCMSHC_EMMC_DLL_CTRL        0x800
>>   #define DWCMSHC_EMMC_DLL_RXCLK        0x804
>> @@ -75,6 +79,11 @@
>>   #define BOUNDARY_OK(addr, len) \
>>       ((addr | (SZ_128M - 1)) == ((addr + len - 1) | (SZ_128M - 1)))
>>   +#define DWCMSHC_SDHCI_CQE_TRNS_MODE    (SDHCI_TRNS_MULTI | \
>> +                     SDHCI_TRNS_BLK_CNT_EN | \
>> +                     SDHCI_TRNS_DMA)
>> +
>> +
>>   enum dwcmshc_rk_type {
>>       DWCMSHC_RK3568,
>>       DWCMSHC_RK3588,
>> @@ -90,7 +99,9 @@ struct rk35xx_priv {
>>     struct dwcmshc_priv {
>>       struct clk    *bus_clk;
>> -    int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA reg */
>> +    int vendor_specific_area1; /* P_VENDOR_SPECIFIC_AREA1 reg */
>> +    int vendor_specific_area2; /* P_VENDOR_SPECIFIC_AREA2 reg */
>> +
>>       void *priv; /* pointer to SoC private stuff */
>>   };
>>   @@ -210,6 +221,90 @@ static void 
>> dwcmshc_hs400_enhanced_strobe(struct mmc_host *mmc,
>>       sdhci_writel(host, vendor, reg);
>>   }
>>   +static int dwcmshc_execute_tuning(struct mmc_host *mmc, u32 opcode)
>> +{
>> +    int err = sdhci_execute_tuning(mmc, opcode);
>> +    struct sdhci_host *host = mmc_priv(mmc);
>> +
>> +    if (err)
>> +        return err;
>> +
>> +    /*
>> +     * Tuning can leave the IP in an active state (Buffer Read 
>> Enable bit
>> +     * set) which prevents the entry to low power states (i.e. 
>> S0i3). Data
>> +     * reset will clear it.
>> +     */
>> +    sdhci_reset(host, SDHCI_RESET_DATA);
>> +
>> +    return 0;
>> +}
>> +
>> +static u32 dwcmshc_cqe_irq_handler(struct sdhci_host *host, u32 
>> intmask)
>> +{
>> +    int cmd_error = 0;
>> +    int data_error = 0;
>> +
>> +    if (!sdhci_cqe_irq(host, intmask, &cmd_error, &data_error))
>> +        return intmask;
>> +
>> +    cqhci_irq(host->mmc, intmask, cmd_error, data_error);
>> +
>> +    return 0;
>> +}
>> +
>> +static void dwcmshc_sdhci_cqe_enable(struct mmc_host *mmc)
>> +{
>> +    struct sdhci_host *host = mmc_priv(mmc);
>> +    u8 ctrl;
>> +
>> +    sdhci_writew(host, DWCMSHC_SDHCI_CQE_TRNS_MODE, 
>> SDHCI_TRANSFER_MODE);
>> +
>> +    sdhci_cqe_enable(mmc);
>> +
>> +    /*
>> +     * The "DesignWare Cores Mobile Storage Host Controller
>> +     * DWC_mshc / DWC_mshc_lite Databook" says:
>> +     * when Host Version 4 Enable" is 1 in Host Control 2 register,
>> +     * SDHCI_CTRL_ADMA32 bit means ADMA2 is selected.
>> +     * Selection of 32-bit/64-bit System Addressing:
>> +     * either 32-bit or 64-bit system addressing is selected by
>> +     * 64-bit Addressing bit in Host Control 2 register.
>> +     *
>> +     * On the other hand the "DesignWare Cores Mobile Storage Host
>> +     * Controller DWC_mshc / DWC_mshc_lite User Guide" says, that we 
>> have to
>> +     * set DMA_SEL to ADMA2 _only_ mode in the Host Control 2 register.
>> +     */
>> +    ctrl = sdhci_readb(host, SDHCI_HOST_CONTROL);
>> +    ctrl &= ~SDHCI_CTRL_DMA_MASK;
>> +    ctrl |= SDHCI_CTRL_ADMA32;
>> +    sdhci_writeb(host, ctrl, SDHCI_HOST_CONTROL);
>> +}
>> +
>> +static void dwcmshc_set_tran_desc(struct cqhci_host *cq_host, u8 
>> **desc,
>> +                  dma_addr_t addr, int len, bool end, bool dma64)
>> +{
>> +    int tmplen, offset;
>> +
>> +    if (likely(!len || BOUNDARY_OK(addr, len))) {
>> +        cqhci_set_tran_desc(*desc, addr, len, end, dma64);
>> +        return;
>> +    }
>> +
>> +    offset = addr & (SZ_128M - 1);
>> +    tmplen = SZ_128M - offset;
>> +    cqhci_set_tran_desc(*desc, addr, tmplen, false, dma64);
>> +
>> +    addr += tmplen;
>> +    len -= tmplen;
>> +    *desc += cq_host->trans_desc_len;
>> +    cqhci_set_tran_desc(*desc, addr, len, end, dma64);
>> +}
>> +
>> +static void dwcmshc_cqhci_dumpregs(struct mmc_host *mmc)
>> +{
>> +    sdhci_dumpregs(mmc_priv(mmc));
>> +}
>> +
>>   static void dwcmshc_rk3568_set_clock(struct sdhci_host *host, 
>> unsigned int clock)
>>   {
>>       struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> @@ -345,6 +440,7 @@ static const struct sdhci_ops sdhci_dwcmshc_ops = {
>>       .get_max_clock        = dwcmshc_get_max_clock,
>>       .reset            = sdhci_reset,
>>       .adma_write_desc    = dwcmshc_adma_write_desc,
>> +    .irq            = dwcmshc_cqe_irq_handler,
>>   };
>>     static const struct sdhci_ops sdhci_dwcmshc_rk35xx_ops = {
>> @@ -379,6 +475,71 @@ static const struct sdhci_pltfm_data 
>> sdhci_dwcmshc_rk35xx_pdata = {
>>              SDHCI_QUIRK2_CLOCK_DIV_ZERO_BROKEN,
>>   };
>>   +static const struct cqhci_host_ops dwcmshc_cqhci_ops = {
>> +    .enable        = dwcmshc_sdhci_cqe_enable,
>> +    .disable    = sdhci_cqe_disable,
>> +    .dumpregs    = dwcmshc_cqhci_dumpregs,
>> +    .set_tran_desc    = dwcmshc_set_tran_desc,
>> +};
>> +
>> +static void dwcmshc_cqhci_init(struct sdhci_host *host, struct 
>> platform_device *pdev)
>> +{
>> +    struct cqhci_host *cq_host;
>> +    struct sdhci_pltfm_host *pltfm_host = sdhci_priv(host);
>> +    struct dwcmshc_priv *priv = sdhci_pltfm_priv(pltfm_host);
>> +    bool dma64 = false;
>> +    u16 clk;
>> +    int err;
>> +
>> +    host->mmc->caps2 |= MMC_CAP2_CQE | MMC_CAP2_CQE_DCMD;
>> +    cq_host = devm_kzalloc(&pdev->dev, sizeof(*cq_host), GFP_KERNEL);
>> +    if (!cq_host) {
>> +        dev_err(mmc_dev(host->mmc), "Unable to setup CQE: not enough 
>> memory\n");
>> +        return;
>> +    }
>> +
>> +    /*
>> +     * For dwcmshc host controller we have to enable internal clock
>> +     * before access to some registers from Vendor Specific Aria 2.
>> +     */
>> +    clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>> +    clk |= SDHCI_CLOCK_INT_EN;
>> +    sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
>> +    clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>> +    if (!(clk & SDHCI_CLOCK_INT_EN)) {
>> +        dev_err(mmc_dev(host->mmc), "Unable to setup CQE: internal 
>> clock enable error\n");
>> +        goto free_cq_host;
>> +    }
>> +
>> +    cq_host->mmio = host->ioaddr + priv->vendor_specific_area2;
>> +    cq_host->ops = &dwcmshc_cqhci_ops;
>> +
>> +    /* Enable using of 128-bit task descriptors */
>> +    dma64 = host->flags & SDHCI_USE_64_BIT_DMA;
>> +    if (dma64) {
>> +        dev_dbg(mmc_dev(host->mmc), "128-bit task descriptors\n");
>> +        cq_host->caps |= CQHCI_TASK_DESC_SZ_128;
>> +    }
>> +    err = cqhci_init(cq_host, host->mmc, dma64);
>> +    if (err) {
>> +        dev_err(mmc_dev(host->mmc), "Unable to setup CQE: error 
>> %d\n", err);
>> +        goto int_clock_disable;
>> +    }
>> +
>> +    dev_dbg(mmc_dev(host->mmc), "CQE init done\n");
>> +
>> +    return;
>> +
>> +int_clock_disable:
>> +    clk = sdhci_readw(host, SDHCI_CLOCK_CONTROL);
>> +    clk &= ~SDHCI_CLOCK_INT_EN;
>> +    sdhci_writew(host, clk, SDHCI_CLOCK_CONTROL);
>> +
>> +free_cq_host:
>> +    devm_kfree(&pdev->dev, cq_host);
>> +}
>> +
>> +
>>   static int dwcmshc_rk35xx_init(struct sdhci_host *host, struct 
>> dwcmshc_priv *dwc_priv)
>>   {
>>       int err;
>> @@ -471,7 +632,7 @@ static int dwcmshc_probe(struct platform_device 
>> *pdev)
>>       struct rk35xx_priv *rk_priv = NULL;
>>       const struct sdhci_pltfm_data *pltfm_data;
>>       int err;
>> -    u32 extra;
>> +    u32 extra, caps;
>>         pltfm_data = device_get_match_data(&pdev->dev);
>>       if (!pltfm_data) {
>> @@ -519,9 +680,12 @@ static int dwcmshc_probe(struct platform_device 
>> *pdev)
>>         priv->vendor_specific_area1 =
>>           sdhci_readl(host, DWCMSHC_P_VENDOR_AREA1) & 
>> DWCMSHC_AREA1_MASK;
>> +    priv->vendor_specific_area2 =
>> +        sdhci_readw(host, DWCMSHC_P_VENDOR_AREA2);
>>         host->mmc_host_ops.request = dwcmshc_request;
>>       host->mmc_host_ops.hs400_enhanced_strobe = 
>> dwcmshc_hs400_enhanced_strobe;
>> +    host->mmc_host_ops.execute_tuning = dwcmshc_execute_tuning;
>>         if (pltfm_data == &sdhci_dwcmshc_rk35xx_pdata) {
>>           rk_priv = devm_kzalloc(&pdev->dev, sizeof(struct 
>> rk35xx_priv), GFP_KERNEL);
>> @@ -547,6 +711,10 @@ static int dwcmshc_probe(struct platform_device 
>> *pdev)
>>           sdhci_enable_v4_mode(host);
>>   #endif
>>   +    caps = sdhci_readl(host, SDHCI_CAPABILITIES);
>> +    if (caps & SDHCI_CAN_64BIT_V4)
>> +        sdhci_enable_v4_mode(host);
>> +
>>       host->mmc->caps |= MMC_CAP_WAIT_WHILE_BUSY;
>>         pm_runtime_get_noresume(dev);
>> @@ -557,6 +725,15 @@ static int dwcmshc_probe(struct platform_device 
>> *pdev)
>>       if (err)
>>           goto err_rpm;
>>   +    /* Setup Command Queue Engine if enabled */
>> +    if (device_property_read_bool(&pdev->dev, "supports-cqe")) {
>> +        if (caps & SDHCI_CAN_64BIT_V4)
>> +            dwcmshc_cqhci_init(host, pdev);
>> +        else
>> +            dev_warn(dev, "Cannot enable CQE without V4 mode 
>> support\n");
>> +    }
>> +
>> +
>>       if (rk_priv)
>>           dwcmshc_rk35xx_postinit(host, priv);