Re: [PATCH 3/6] PCI: imx6: add IMX8MM support

From: Bjorn Helgaas <helgaas@kernel.org>
To: Tim Harvey <tharvey@gateworks.com>
Cc: "Richard Zhu" <hongxing.zhu@nxp.com>,
	"Lucas Stach" <l.stach@pengutronix.de>,
	"Bjorn Helgaas" <bhelgaas@google.com>,
	"Rob Herring" <robh+dt@kernel.org>,
	"Shawn Guo" <shawnguo@kernel.org>,
	"Sascha Hauer" <s.hauer@pengutronix.de>,
	"Pengutronix Kernel Team" <kernel@pengutronix.de>,
	"Fabio Estevam" <festevam@gmail.com>,
	"NXP Linux Team" <linux-imx@nxp.com>,
	linux-pci@vger.kernel.org, linux-arm-kernel@lists.infradead.org,
	devicetree@vger.kernel.org, linux-kernel@vger.kernel.org,
	"Krzysztof Wilczyński" <kw@linux.com>,
	"Lorenzo Pieralisi" <lorenzo.pieralisi@arm.com>
Subject: Re: [PATCH 3/6] PCI: imx6: add IMX8MM support
Date: Thu, 29 Jul 2021 17:29:17 -0500	[thread overview]
Message-ID: <20210729222917.GA998237@bjorn-Precision-5520> (raw)
In-Reply-To: <20210723204958.7186-4-tharvey@gateworks.com>

Capitalize first letter of subject to match history:

  PCI: imx6: Add IMX8MM support

On Fri, Jul 23, 2021 at 01:49:55PM -0700, Tim Harvey wrote:
> Add IMX8MM support to the existing driver which shares most
> functionality with the IMX8MM.
> 
> Signed-off-by: Tim Harvey <tharvey@gateworks.com>
> ---
>  drivers/pci/controller/dwc/pci-imx6.c | 103 +++++++++++++++++++++++++-
>  1 file changed, 102 insertions(+), 1 deletion(-)

> +	case IMX8MM:
> +		offset = imx6_pcie_grp_offset(imx6_pcie);
> +
> +		dev_info(imx6_pcie->pci->dev, "%s REF_CLK is used!.\n",
> +				imx6_pcie->ext_osc ? "EXT" : "PLL");
> +		if (imx6_pcie->ext_osc) {
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MQ_GPR_PCIE_REF_USE_PAD, 0);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_REF_CLK_SEL,
> +					IMX8MM_GPR_PCIE_REF_CLK_SEL);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_AUX_EN,
> +					IMX8MM_GPR_PCIE_AUX_EN);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_POWER_OFF, 0);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_SSC_EN, 0);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_REF_CLK_SEL,
> +					IMX8MM_GPR_PCIE_REF_CLK_EXT);
> +			udelay(100);
> +			/* Do the PHY common block reset */
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_CMN_RST,
> +					IMX8MM_GPR_PCIE_CMN_RST);
> +			udelay(200);
> +		} else {
> +			/* Configure the internal PLL as REF clock */
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MQ_GPR_PCIE_REF_USE_PAD, 0);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_REF_CLK_SEL,
> +					IMX8MM_GPR_PCIE_REF_CLK_SEL);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_AUX_EN,
> +					IMX8MM_GPR_PCIE_AUX_EN);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_POWER_OFF, 0);
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_SSC_EN, 0);

Seems like all the above is common to both cases?  If so, it's a shame
to repeat it because it makes it hard to see what's different.

> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_REF_CLK_SEL,
> +					IMX8MM_GPR_PCIE_REF_CLK_PLL);
> +			udelay(100);
> +			/* Configure the PHY */
> +			writel(PCIE_PHY_CMN_REG62_PLL_CLK_OUT,
> +					imx6_pcie->phy_base + PCIE_PHY_CMN_REG62);
> +			writel(PCIE_PHY_CMN_REG64_AUX_RX_TX_TERM,
> +					imx6_pcie->phy_base + PCIE_PHY_CMN_REG64);
> +			/* Do the PHY common block reset */
> +			regmap_update_bits(imx6_pcie->iomuxc_gpr, offset,
> +					IMX8MM_GPR_PCIE_CMN_RST,
> +					IMX8MM_GPR_PCIE_CMN_RST);
> +			udelay(200);
> +		}
> +		/*

> +		 * In order to pass the compliance tests.
> +		 * Configure the TRSV regiser of iMX8MM PCIe PHY.

The "In order to ..." line isn't quite a sentence.  Maybe it should be
joined with the second line?

s/regiser/register/ ?

> +		 */
> +		writel(PCIE_PHY_TRSV_REG5_GEN1_DEEMP,
> +				imx6_pcie->phy_base + PCIE_PHY_TRSV_REG5);

> +	/* check for EXT osc */

Since you have a comment here, it would be useful to spell out "osc"
for newbies like me.  I assume it's short for "oscillator"?

> +	imx6_pcie->ext_osc = of_property_read_bool(node, "fsl,ext-osc");