| lists.openwall.net | lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC | |
|
Open Source and information security mailing list archives
| ||
|
Message-ID: <20191226201146.GA1803@voidbox>
Date: Thu, 26 Dec 2019 21:11:46 +0100
From: Remi Pommarel <repk@...plefau.lt>
To: Jerome Brunet <jbrunet@...libre.com>
Cc: Kishon Vijay Abraham I <kishon@...com>,
Yue Wang <yue.wang@...ogic.com>,
Lorenzo Pieralisi <lorenzo.pieralisi@....com>,
Andrew Murray <andrew.murray@....com>,
Kevin Hilman <khilman@...libre.com>,
Martin Blumenstingl <martin.blumenstingl@...glemail.com>,
Neil Armstrong <narmstrong@...libre.com>,
linux-kernel@...r.kernel.org, linux-pci@...r.kernel.org,
linux-amlogic@...ts.infradead.org
Subject: Re: [PATCH v2 1/3] phy: amlogic: Add Amlogic AXG MIPI/PCIE PHY Driver
On Thu, Dec 26, 2019 at 10:39:36AM +0100, Jerome Brunet wrote:
>
> On Mon 23 Dec 2019 at 22:45, Remi Pommarel <repk@...plefau.lt> wrote:
>
> > This adds support for the MIPI PHY also needed for PCIE found in the
> > Amlogic AXG SoC Family.
> >
> > MIPI or PCIE selection is done by the #phy-cells, making the mode
> > static and exclusive.
> >
> > For now only PCIE fonctionality is supported.
> >
> > This PHY will be used to replace the mipi_enable clock gating logic
> > which was mistakenly added in the clock subsystem. This also activate
> > a non documented band gap bit in those registers that allows reliable
> > PCIE clock signal generation on AXG platforms.
> >
> > Signed-off-by: Remi Pommarel <repk@...plefau.lt>
> > ---
> > drivers/phy/amlogic/Kconfig | 11 ++
> > drivers/phy/amlogic/Makefile | 1 +
> > drivers/phy/amlogic/phy-meson-axg-mipi-pcie.c | 176 ++++++++++++++++++
> > 3 files changed, 188 insertions(+)
> > create mode 100644 drivers/phy/amlogic/phy-meson-axg-mipi-pcie.c
> >
> > diff --git a/drivers/phy/amlogic/Kconfig b/drivers/phy/amlogic/Kconfig
> > index af774ac2b934..1eeb75d018e3 100644
> > --- a/drivers/phy/amlogic/Kconfig
> > +++ b/drivers/phy/amlogic/Kconfig
> > @@ -59,3 +59,14 @@ config PHY_MESON_G12A_USB3_PCIE
> > Enable this to support the Meson USB3 + PCIE Combo PHY found
> > in Meson G12A SoCs.
> > If unsure, say N.
> > +
> > +config PHY_MESON_AXG_MIPI_PCIE
> > + tristate "Meson AXG MIPI + PCIE PHY driver"
> > + default ARCH_MESON
> > + depends on OF && (ARCH_MESON || COMPILE_TEST)
> > + select GENERIC_PHY
> > + select MFD_SYSCON
> > + help
> > + Enable this to support the Meson MIPI + PCIE PHY found
> > + in Meson AXG SoCs.
> > + If unsure, say N.
> > diff --git a/drivers/phy/amlogic/Makefile b/drivers/phy/amlogic/Makefile
> > index 11d1c42ac2be..2167330a0ae8 100644
> > --- a/drivers/phy/amlogic/Makefile
> > +++ b/drivers/phy/amlogic/Makefile
> > @@ -4,3 +4,4 @@ obj-$(CONFIG_PHY_MESON_GXL_USB2) += phy-meson-gxl-usb2.o
> > obj-$(CONFIG_PHY_MESON_G12A_USB2) += phy-meson-g12a-usb2.o
> > obj-$(CONFIG_PHY_MESON_GXL_USB3) += phy-meson-gxl-usb3.o
> > obj-$(CONFIG_PHY_MESON_G12A_USB3_PCIE) += phy-meson-g12a-usb3-pcie.o
> > +obj-$(CONFIG_PHY_MESON_AXG_MIPI_PCIE) += phy-meson-axg-mipi-pcie.o
> > diff --git a/drivers/phy/amlogic/phy-meson-axg-mipi-pcie.c b/drivers/phy/amlogic/phy-meson-axg-mipi-pcie.c
> > new file mode 100644
> > index 000000000000..006aa8cdfc47
> > --- /dev/null
> > +++ b/drivers/phy/amlogic/phy-meson-axg-mipi-pcie.c
> > @@ -0,0 +1,176 @@
> > +// SPDX-License-Identifier: GPL-2.0
> > +/*
> > + * Amlogic AXG MIPI + PCIE PHY driver
> > + *
> > + * Copyright (C) 2019 Remi Pommarel <repk@...plefau.lt>
> > + */
> > +#include <linux/module.h>
> > +#include <linux/phy/phy.h>
> > +#include <linux/regmap.h>
> > +#include <linux/mfd/syscon.h>
> > +#include <linux/platform_device.h>
> > +#include <dt-bindings/phy/phy.h>
> > +
> > +#define HHI_MIPI_CNTL0 0x00
> > +#define HHI_MIPI_CNTL0_COMMON_BLOCK GENMASK(31, 28)
> > +#define HHI_MIPI_CNTL0_ENABLE BIT(29)
> > +#define HHI_MIPI_CNTL0_BANDGAP BIT(26)
> > +#define HHI_MIPI_CNTL0_DECODE_TO_RTERM GENMASK(15, 12)
> > +#define HHI_MIPI_CNTL0_OUTPUT_EN BIT(3)
> > +
> > +#define HHI_MIPI_CNTL1 0x01
> > +#define HHI_MIPI_CNTL1_CH0_CML_PDR_EN BIT(12)
> > +#define HHI_MIPI_CNTL1_LP_ABILITY GENMASK(5, 4)
> > +#define HHI_MIPI_CNTL1_LP_RESISTER BIT(3)
> > +#define HHI_MIPI_CNTL1_INPUT_SETTING BIT(2)
> > +#define HHI_MIPI_CNTL1_INPUT_SEL BIT(1)
> > +#define HHI_MIPI_CNTL1_PRBS7_EN BIT(0)
> > +
> > +#define HHI_MIPI_CNTL2 0x02
> > +#define HHI_MIPI_CNTL2_CH_PU GENMASK(31, 25)
> > +#define HHI_MIPI_CNTL2_CH_CTL GENMASK(24, 19)
> > +#define HHI_MIPI_CNTL2_CH0_DIGDR_EN BIT(18)
> > +#define HHI_MIPI_CNTL2_CH_DIGDR_EN BIT(17)
> > +#define HHI_MIPI_CNTL2_LPULPS_EN BIT(16)
> > +#define HHI_MIPI_CNTL2_CH_EN(n) BIT(15 - (n))
> > +#define HHI_MIPI_CNTL2_CH0_LP_CTL GENMASK(10, 1)
> > +
> > +struct phy_axg_mipi_pcie_priv {
> > + struct phy *phy;
> > + unsigned int mode;
> > + struct regmap *regmap;
> > +};
> > +
> > +static const struct regmap_config phy_axg_mipi_pcie_regmap_conf = {
> > + .reg_bits = 8,
> > + .val_bits = 32,
> > + .reg_stride = 4,
> > + .max_register = HHI_MIPI_CNTL2,
> > +};
> > +
> > +static int phy_axg_mipi_pcie_power_on(struct phy *phy)
> > +{
> > + struct phy_axg_mipi_pcie_priv *priv = phy_get_drvdata(phy);
> > +
> > + /* MIPI not supported yet */
> > + if (priv->mode != PHY_TYPE_PCIE)
> > + return 0;
> > +
> > + regmap_update_bits(priv->regmap, HHI_MIPI_CNTL0,
> > + HHI_MIPI_CNTL0_BANDGAP, HHI_MIPI_CNTL0_BANDGAP);
> > +
> > + regmap_update_bits(priv->regmap, HHI_MIPI_CNTL0,
> > + HHI_MIPI_CNTL0_ENABLE, HHI_MIPI_CNTL0_ENABLE);
> > + return 0;
> > +}
> > +
> > +static int phy_axg_mipi_pcie_power_off(struct phy *phy)
> > +{
> > + struct phy_axg_mipi_pcie_priv *priv = phy_get_drvdata(phy);
> > +
> > + /* MIPI not supported yet */
> > + if (priv->mode != PHY_TYPE_PCIE)
> > + return 0;
> > +
> > + regmap_update_bits(priv->regmap, HHI_MIPI_CNTL0,
> > + HHI_MIPI_CNTL0_BANDGAP, 0);
> > + regmap_update_bits(priv->regmap, HHI_MIPI_CNTL0,
> > + HHI_MIPI_CNTL0_ENABLE, 0);
> > + return 0;
> > +}
> > +
> > +static int phy_axg_mipi_pcie_init(struct phy *phy)
> > +{
> > + return 0;
> > +}
> > +
> > +static int phy_axg_mipi_pcie_exit(struct phy *phy)
> > +{
> > + return 0;
> > +}
> > +
> > +static const struct phy_ops phy_axg_mipi_pcie_ops = {
> > + .init = phy_axg_mipi_pcie_init,
> > + .exit = phy_axg_mipi_pcie_exit,
> > + .power_on = phy_axg_mipi_pcie_power_on,
> > + .power_off = phy_axg_mipi_pcie_power_off,
> > + .owner = THIS_MODULE,
> > +};
> > +
> > +static struct phy *phy_axg_mipi_pcie_xlate(struct device *dev,
> > + struct of_phandle_args *args)
> > +{
> > + struct phy_axg_mipi_pcie_priv *priv = dev_get_drvdata(dev);
> > + unsigned int mode;
> > +
> > + if (args->args_count != 1) {
> > + dev_err(dev, "invalid number of arguments\n");
> > + return ERR_PTR(-EINVAL);
> > + }
> > +
> > + mode = args->args[0];
> > +
> > + /* MIPI mode is not supported yet */
> > + if (mode != PHY_TYPE_PCIE) {
> > + dev_err(dev, "invalid phy mode select argument\n");
> > + return ERR_PTR(-EINVAL);
> > + }
> > +
> > + priv->mode = mode;
> > + return priv->phy;
> > +}
> > +
> > +static int phy_axg_mipi_pcie_probe(struct platform_device *pdev)
> > +{
> > + struct phy_provider *pphy;
> > + struct device *dev = &pdev->dev;
> > + struct phy_axg_mipi_pcie_priv *priv;
> > + struct device_node *np = dev->of_node;
> > + int ret;
> > +
> > + priv = devm_kmalloc(dev, sizeof(*priv), GFP_KERNEL);
> > + if (!priv)
> > + return -ENOMEM;
> > +
> > + /* Get the hhi system controller node */
> > + priv->regmap = syscon_node_to_regmap(of_get_parent(dev->of_node));
> > + if (IS_ERR(priv->regmap)) {
> > + dev_err(dev, "failed to get HHI regmap\n");
> > + return PTR_ERR(priv->regmap);
> > + }
>
> Remi,
>
> Unless we are absolutely sure this will be *AXG ONLY*, I would
> prefer if you get the registers without the syscon.
>
> Having it introduce some kind of dependency between the 2 which is
> likely to make this driver SoC specific.
>
> It was clearly wrong for the clock controller to map these regiters, and
> if there is a possibility that this driver is used on other SoCs, I
> would prefer if we did not carry that mistake over. I would prefer if we
> fixed the clock controller so you don't need syscon here.
Jerome thank you for reviewing this,
Sure I will remove access to the registers through syscon system. Just
to be sure we are on the same page here. What you suggest is keeping the
two PHYs approach from this patchset (instead of the one PHY in v3),
even if there is not exactly two PHYs, right ?
Thanks,
--
Remi
>
> > +
> > + priv->phy = devm_phy_create(dev, np, &phy_axg_mipi_pcie_ops);
> > + if (IS_ERR(priv->phy)) {
> > + ret = PTR_ERR(priv->phy);
> > + if (ret != -EPROBE_DEFER)
> > + dev_err(dev, "failed to create PHY\n");
> > + return ret;
> > + }
> > +
> > + phy_set_drvdata(priv->phy, priv);
> > + dev_set_drvdata(dev, priv);
> > + pphy = devm_of_phy_provider_register(dev, phy_axg_mipi_pcie_xlate);
> > +
> > + return PTR_ERR_OR_ZERO(pphy);
> > +}
> > +
> > +static const struct of_device_id phy_axg_mipi_pcie_of_match[] = {
> > + {
> > + .compatible = "amlogic,axg-mipi-pcie-phy",
> > + },
> > + { },
> > +};
> > +MODULE_DEVICE_TABLE(of, phy_axg_mipi_pcie_of_match);
> > +
> > +static struct platform_driver phy_axg_mipi_pcie_driver = {
> > + .probe = phy_axg_mipi_pcie_probe,
> > + .driver = {
> > + .name = "phy-axg-mipi-pcie",
> > + .of_match_table = phy_axg_mipi_pcie_of_match,
> > + },
> > +};
> > +module_platform_driver(phy_axg_mipi_pcie_driver);
> > +
> > +MODULE_AUTHOR("Remi Pommarel <repk@...plefau.lt>");
> > +MODULE_DESCRIPTION("Amlogic AXG MIPI + PCIE PHY driver");
> > +MODULE_LICENSE("GPL v2");
>
Powered by blists - more mailing lists