[PATCH 0/3] Fixes for "PCIE support for i.MX8MQ"

[PATCH 0/3] Fixes for "PCIE support for i.MX8MQ"

[Andrey Smirnov]

Lorenzo:

These are patches fixing things pointed out by Bjorn in [1] and [2] as
well as something I noticed while giving pci/next branch a try (#endif
placement). Hopefully this is at least somewhat helpful.

Thanks,
Andrey Smirnov

[1] lkml.kernel.org/r/20181214203828.GD20725@google.com
[2] lkml.kernel.org/r/20181214203042.GC20725@google.com

Andrey Smirnov (3):
  PCI: imx6: Fixup Kconfig and #endif placement
  PCI: imx6: Invert checks in imx6_pcie_reset_phy() and
    imx6_setup_phy_mpll()
  PCI: imx6: Make fallthrough comments more consistent

 drivers/pci/controller/dwc/Kconfig    |  4 ++--
 drivers/pci/controller/dwc/pci-imx6.c | 29 ++++++++++++++++++---------
 2 files changed, 21 insertions(+), 12 deletions(-)

-- 
2.19.1

[PATCH 1/3] PCI: imx6: Fixup Kconfig and #endif placement

[Andrey Smirnov]

Fixup Kconfig, which lost SOC_IMX7D during merge as well as move
"endif" to its original placement to only imx6q_pcie_abort_handler()
when building for AArch64

Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Fabio Estevam <fabio.estevam@nxp.com>
Cc: Chris Healy <cphealy@gmail.com>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Leonard Crestez <leonard.crestez@nxp.com>
Cc: "A.s. Dong" <aisheng.dong@nxp.com>
Cc: Richard Zhu <hongxing.zhu@nxp.com>
Cc: linux-imx@nxp.com
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-pci@vger.kernel.org
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
 drivers/pci/controller/dwc/Kconfig    | 4 ++--
 drivers/pci/controller/dwc/pci-imx6.c | 2 +-
 2 files changed, 3 insertions(+), 3 deletions(-)

diff --git a/drivers/pci/controller/dwc/Kconfig b/drivers/pci/controller/dwc/Kconfig
index 0c8142529241..b5202ed78762 100644
--- a/drivers/pci/controller/dwc/Kconfig
+++ b/drivers/pci/controller/dwc/Kconfig
@@ -89,8 +89,8 @@ config PCI_EXYNOS
 	select PCIE_DW_HOST
 
 config PCI_IMX6
-	bool "Freescale i.MX6/7 PCIe controller"
-	depends on SOC_IMX8MQ || SOC_IMX6Q || (ARM && COMPILE_TEST)
+	bool "Freescale i.MX6/7/8 PCIe controller"
+	depends on SOC_IMX8MQ || SOC_IMX6Q || SOC_IMX7D || (ARM && COMPILE_TEST)
 	depends on PCI_MSI_IRQ_DOMAIN
 	select PCIE_DW_HOST
 
diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
index e563ca9bdb78..4b7f638b8aff 100644
--- a/drivers/pci/controller/dwc/pci-imx6.c
+++ b/drivers/pci/controller/dwc/pci-imx6.c
@@ -311,6 +311,7 @@ static int imx6q_pcie_abort_handler(unsigned long addr,
 
 	return 1;
 }
+#endif
 
 static int imx6_pcie_attach_pd(struct device *dev)
 {
@@ -389,7 +390,6 @@ static void imx6_pcie_assert_core_reset(struct imx6_pcie *imx6_pcie)
 				ret);
 	}
 }
-#endif
 
 static int imx6_pcie_enable_ref_clk(struct imx6_pcie *imx6_pcie)
 {
-- 
2.19.1

[PATCH 2/3] PCI: imx6: Invert checks in imx6_pcie_reset_phy() and imx6_setup_phy_mpll()

[Andrey Smirnov]

In order to avoid having potentially ever growing list of variants
that don't support methods implemented in imx6_pcie_reset_phy() and
imx6_setup_phy_mpll(), change logical checks in the to check for SoC's
that _do_ support what they implement. While at it, share the code via
a small helper function.

Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Fabio Estevam <fabio.estevam@nxp.com>
Cc: Chris Healy <cphealy@gmail.com>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Leonard Crestez <leonard.crestez@nxp.com>
Cc: "A.s. Dong" <aisheng.dong@nxp.com>
Cc: Richard Zhu <hongxing.zhu@nxp.com>
Cc: linux-imx@nxp.com
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-pci@vger.kernel.org
Suggested-by: Bjorn Helgaas <helgaas@kernel.org>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
 drivers/pci/controller/dwc/pci-imx6.c | 13 +++++++++----
 1 file changed, 9 insertions(+), 4 deletions(-)

diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
index 4b7f638b8aff..59658577e81d 100644
--- a/drivers/pci/controller/dwc/pci-imx6.c
+++ b/drivers/pci/controller/dwc/pci-imx6.c
@@ -256,12 +256,18 @@ static int pcie_phy_write(struct imx6_pcie *imx6_pcie, int addr, int data)
 	return 0;
 }
 
+static bool imx6_pcie_has_imx6_phy(struct imx6_pcie *imx6_pcie)
+{
+	return imx6_pcie->variant == IMX6Q ||
+	       imx6_pcie->variant == IMX6SX ||
+	       imx6_pcie->variant == IMX6QP;
+}
+
 static void imx6_pcie_reset_phy(struct imx6_pcie *imx6_pcie)
 {
 	u32 tmp;
 
-	if (imx6_pcie->variant == IMX7D ||
-	    imx6_pcie->variant == IMX8MQ)
+	if (!imx6_pcie_has_imx6_phy(imx6_pcie))
 		return;
 
 	pcie_phy_read(imx6_pcie, PHY_RX_OVRD_IN_LO, &tmp);
@@ -637,8 +643,7 @@ static int imx6_setup_phy_mpll(struct imx6_pcie *imx6_pcie)
 	int mult, div;
 	u32 val;
 
-	if (imx6_pcie->variant == IMX7D ||
-	    imx6_pcie->variant == IMX8MQ)
+	if (!imx6_pcie_has_imx6_phy(imx6_pcie))
 		return 0;
 
 	switch (phy_rate) {
-- 
2.19.1

[PATCH 3/3] PCI: imx6: Make fallthrough comments more consistent

[Andrey Smirnov]

Convert all fallthrough comments to say "fall through", as well as
modify their placement to the point where the "break" would normally
be.

Cc: Bjorn Helgaas <bhelgaas@google.com>
Cc: Fabio Estevam <fabio.estevam@nxp.com>
Cc: Chris Healy <cphealy@gmail.com>
Cc: Lucas Stach <l.stach@pengutronix.de>
Cc: Leonard Crestez <leonard.crestez@nxp.com>
Cc: "A.s. Dong" <aisheng.dong@nxp.com>
Cc: Richard Zhu <hongxing.zhu@nxp.com>
Cc: linux-imx@nxp.com
Cc: linux-arm-kernel@lists.infradead.org
Cc: linux-kernel@vger.kernel.org
Cc: linux-pci@vger.kernel.org
Suggested-by: Bjorn Helgaas <helgaas@kernel.org>
Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
---
 drivers/pci/controller/dwc/pci-imx6.c | 14 +++++++++-----
 1 file changed, 9 insertions(+), 5 deletions(-)

diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
index 59658577e81d..a0510e185d44 100644
--- a/drivers/pci/controller/dwc/pci-imx6.c
+++ b/drivers/pci/controller/dwc/pci-imx6.c
@@ -362,7 +362,8 @@ static void imx6_pcie_assert_core_reset(struct imx6_pcie *imx6_pcie)
 
 	switch (imx6_pcie->variant) {
 	case IMX7D:
-	case IMX8MQ: /* FALLTHROUGH */
+		/* fall through */
+	case IMX8MQ:
 		reset_control_assert(imx6_pcie->pciephy_reset);
 		reset_control_assert(imx6_pcie->apps_reset);
 		break;
@@ -415,7 +416,8 @@ static int imx6_pcie_enable_ref_clk(struct imx6_pcie *imx6_pcie)
 		regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12,
 				   IMX6SX_GPR12_PCIE_TEST_POWERDOWN, 0);
 		break;
-	case IMX6QP:		/* FALLTHROUGH */
+	case IMX6QP:
+		/* fall through */
 	case IMX6Q:
 		/* power up core phy and enable ref clock */
 		regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR1,
@@ -607,7 +609,7 @@ static void imx6_pcie_init_phy(struct imx6_pcie *imx6_pcie)
 		regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12,
 				   IMX6SX_GPR12_PCIE_RX_EQ_MASK,
 				   IMX6SX_GPR12_PCIE_RX_EQ_2);
-		/* FALLTHROUGH */
+		/* fall through */
 	default:
 		regmap_update_bits(imx6_pcie->iomuxc_gpr, IOMUXC_GPR12,
 				   IMX6Q_GPR12_PCIE_CTL_2, 0 << 10);
@@ -731,7 +733,8 @@ static void imx6_pcie_ltssm_enable(struct device *dev)
 				   IMX6Q_GPR12_PCIE_CTL_2);
 		break;
 	case IMX7D:
-	case IMX8MQ:		/* FALLTHROUGH */
+		/* fall through */
+	case IMX8MQ:
 		reset_control_deassert(imx6_pcie->apps_reset);
 		break;
 	}
@@ -1076,7 +1079,8 @@ static int imx6_pcie_probe(struct platform_device *pdev)
 		}
 		break;
 	case IMX8MQ:
-	case IMX7D:		/* FALLTHROUGH */
+		/* fall through */
+	case IMX7D:
 		imx6_pcie->pciephy_reset = devm_reset_control_get_exclusive(dev,
 									    "pciephy");
 		if (IS_ERR(imx6_pcie->pciephy_reset)) {
-- 
2.19.1

Re: [PATCH 2/3] PCI: imx6: Invert checks in imx6_pcie_reset_phy() and imx6_setup_phy_mpll()

[Leonard Crestez]

On 12/17/2018 1:09 AM, Andrey Smirnov wrote:
> In order to avoid having potentially ever growing list of variants
> that don't support methods implemented in imx6_pcie_reset_phy() and
> imx6_setup_phy_mpll(), change logical checks in the to check for SoC's
> that _do_ support what they implement. While at it, share the code via
> a small helper function.
>   
> +static bool imx6_pcie_has_imx6_phy(struct imx6_pcie *imx6_pcie)
> +{
> +	return imx6_pcie->variant == IMX6Q ||
> +	       imx6_pcie->variant == IMX6SX ||
> +	       imx6_pcie->variant == IMX6QP;
> +}

This would be an ideal match for adding a field inside drvdata, sadly 
that was part of a series which stalled:

https://patchwork.kernel.org/patch/10712261/

Re: [PATCH 0/3] Fixes for "PCIE support for i.MX8MQ"

[Lorenzo Pieralisi]

On Sun, Dec 16, 2018 at 03:09:13PM -0800, Andrey Smirnov wrote:
> Lorenzo:
> 
> These are patches fixing things pointed out by Bjorn in [1] and [2] as
> well as something I noticed while giving pci/next branch a try (#endif
> placement). Hopefully this is at least somewhat helpful.
> 
> Thanks,
> Andrey Smirnov
> 
> [1] lkml.kernel.org/r/20181214203828.GD20725@google.com
> [2] lkml.kernel.org/r/20181214203042.GC20725@google.com
> 
> Andrey Smirnov (3):
>   PCI: imx6: Fixup Kconfig and #endif placement
>   PCI: imx6: Invert checks in imx6_pcie_reset_phy() and
>     imx6_setup_phy_mpll()
>   PCI: imx6: Make fallthrough comments more consistent
> 
>  drivers/pci/controller/dwc/Kconfig    |  4 ++--
>  drivers/pci/controller/dwc/pci-imx6.c | 29 ++++++++++++++++++---------
>  2 files changed, 21 insertions(+), 12 deletions(-)

Andrey,

please follow-up Lucas' remarks and submit a self-consistent patchset
that I can apply on top of my pci/dwc branch (where I removed your
previously queued patches).

Thanks,
Lorenzo

Re: [PATCH 3/3] PCI: imx6: Make fallthrough comments more consistent

[Bjorn Helgaas]

On Sun, Dec 16, 2018 at 03:09:16PM -0800, Andrey Smirnov wrote:
> Convert all fallthrough comments to say "fall through", as well as
> modify their placement to the point where the "break" would normally
> be.
> 
> Cc: Bjorn Helgaas <bhelgaas@google.com>
> Cc: Fabio Estevam <fabio.estevam@nxp.com>
> Cc: Chris Healy <cphealy@gmail.com>
> Cc: Lucas Stach <l.stach@pengutronix.de>
> Cc: Leonard Crestez <leonard.crestez@nxp.com>
> Cc: "A.s. Dong" <aisheng.dong@nxp.com>
> Cc: Richard Zhu <hongxing.zhu@nxp.com>
> Cc: linux-imx@nxp.com
> Cc: linux-arm-kernel@lists.infradead.org
> Cc: linux-kernel@vger.kernel.org
> Cc: linux-pci@vger.kernel.org
> Suggested-by: Bjorn Helgaas <helgaas@kernel.org>

I didn't make it very clear, but my suggestion was really to remove
the annotation completely; see below.

> Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> ---
>  drivers/pci/controller/dwc/pci-imx6.c | 14 +++++++++-----
>  1 file changed, 9 insertions(+), 5 deletions(-)
> 
> diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
> index 59658577e81d..a0510e185d44 100644
> --- a/drivers/pci/controller/dwc/pci-imx6.c
> +++ b/drivers/pci/controller/dwc/pci-imx6.c
> @@ -362,7 +362,8 @@ static void imx6_pcie_assert_core_reset(struct imx6_pcie *imx6_pcie)
>  
>  	switch (imx6_pcie->variant) {
>  	case IMX7D:
> -	case IMX8MQ: /* FALLTHROUGH */
> +		/* fall through */
> +	case IMX8MQ:
>  		reset_control_assert(imx6_pcie->pciephy_reset);
>  		reset_control_assert(imx6_pcie->apps_reset);
>  		break;

IMO this use of "fall through" is superfluous and unusual in the Linux
source.

A "fall through" comment would be useful if the IMX7D case had
executable code but no "break".  Then the comment shows that the
intent was to execute *both* the IMX7D code and the IMX8MQ code and
the lack of a "break" was intentional.

In this case, the intent is to treat IMX7D and IMX8MQ the same, and
there's no executable code specifically for the IMX7D.  I think it's
easiest to read that when the list of identical cases is all together
without the comment in the middle, i.e., as

>  	case IMX7D:
> 	case IMX8MQ:
>  		reset_control_assert(imx6_pcie->pciephy_reset);

rather than this:

>  	case IMX7D:
> 		/* fall through */
> 	case IMX8MQ:
>  		reset_control_assert(imx6_pcie->pciephy_reset);

Re: [PATCH 2/3] PCI: imx6: Invert checks in imx6_pcie_reset_phy() and imx6_setup_phy_mpll()

[Andrey Smirnov]

On Mon, Dec 17, 2018 at 2:24 AM Leonard Crestez <leonard.crestez@nxp.com> wrote:
>
> On 12/17/2018 1:09 AM, Andrey Smirnov wrote:
> > In order to avoid having potentially ever growing list of variants
> > that don't support methods implemented in imx6_pcie_reset_phy() and
> > imx6_setup_phy_mpll(), change logical checks in the to check for SoC's
> > that _do_ support what they implement. While at it, share the code via
> > a small helper function.
> >
> > +static bool imx6_pcie_has_imx6_phy(struct imx6_pcie *imx6_pcie)
> > +{
> > +     return imx6_pcie->variant == IMX6Q ||
> > +            imx6_pcie->variant == IMX6SX ||
> > +            imx6_pcie->variant == IMX6QP;
> > +}
>
> This would be an ideal match for adding a field inside drvdata, sadly
> that was part of a series which stalled:
>
> https://patchwork.kernel.org/patch/10712261/

OK, I can pick up that individual patch into this series.

Thanks,
Andrey Smirnov

Re: [PATCH 0/3] Fixes for "PCIE support for i.MX8MQ"

[Andrey Smirnov]

On Mon, Dec 17, 2018 at 3:01 AM Lorenzo Pieralisi
<lorenzo.pieralisi@arm.com> wrote:
>
> On Sun, Dec 16, 2018 at 03:09:13PM -0800, Andrey Smirnov wrote:
> > Lorenzo:
> >
> > These are patches fixing things pointed out by Bjorn in [1] and [2] as
> > well as something I noticed while giving pci/next branch a try (#endif
> > placement). Hopefully this is at least somewhat helpful.
> >
> > Thanks,
> > Andrey Smirnov
> >
> > [1] lkml.kernel.org/r/20181214203828.GD20725@google.com
> > [2] lkml.kernel.org/r/20181214203042.GC20725@google.com
> >
> > Andrey Smirnov (3):
> >   PCI: imx6: Fixup Kconfig and #endif placement
> >   PCI: imx6: Invert checks in imx6_pcie_reset_phy() and
> >     imx6_setup_phy_mpll()
> >   PCI: imx6: Make fallthrough comments more consistent
> >
> >  drivers/pci/controller/dwc/Kconfig    |  4 ++--
> >  drivers/pci/controller/dwc/pci-imx6.c | 29 ++++++++++++++++++---------
> >  2 files changed, 21 insertions(+), 12 deletions(-)
>
> Andrey,
>
> please follow-up Lucas' remarks and submit a self-consistent patchset
> that I can apply on top of my pci/dwc branch (where I removed your
> previously queued patches).
>

Sure, will do.

Thanks,
Andrey Smirnov

Re: [PATCH 3/3] PCI: imx6: Make fallthrough comments more consistent

[Andrey Smirnov]

On Mon, Dec 17, 2018 at 5:58 AM Bjorn Helgaas <helgaas@kernel.org> wrote:
>
> On Sun, Dec 16, 2018 at 03:09:16PM -0800, Andrey Smirnov wrote:
> > Convert all fallthrough comments to say "fall through", as well as
> > modify their placement to the point where the "break" would normally
> > be.
> >
> > Cc: Bjorn Helgaas <bhelgaas@google.com>
> > Cc: Fabio Estevam <fabio.estevam@nxp.com>
> > Cc: Chris Healy <cphealy@gmail.com>
> > Cc: Lucas Stach <l.stach@pengutronix.de>
> > Cc: Leonard Crestez <leonard.crestez@nxp.com>
> > Cc: "A.s. Dong" <aisheng.dong@nxp.com>
> > Cc: Richard Zhu <hongxing.zhu@nxp.com>
> > Cc: linux-imx@nxp.com
> > Cc: linux-arm-kernel@lists.infradead.org
> > Cc: linux-kernel@vger.kernel.org
> > Cc: linux-pci@vger.kernel.org
> > Suggested-by: Bjorn Helgaas <helgaas@kernel.org>
>
> I didn't make it very clear, but my suggestion was really to remove
> the annotation completely; see below.
>
> > Signed-off-by: Andrey Smirnov <andrew.smirnov@gmail.com>
> > ---
> >  drivers/pci/controller/dwc/pci-imx6.c | 14 +++++++++-----
> >  1 file changed, 9 insertions(+), 5 deletions(-)
> >
> > diff --git a/drivers/pci/controller/dwc/pci-imx6.c b/drivers/pci/controller/dwc/pci-imx6.c
> > index 59658577e81d..a0510e185d44 100644
> > --- a/drivers/pci/controller/dwc/pci-imx6.c
> > +++ b/drivers/pci/controller/dwc/pci-imx6.c
> > @@ -362,7 +362,8 @@ static void imx6_pcie_assert_core_reset(struct imx6_pcie *imx6_pcie)
> >
> >       switch (imx6_pcie->variant) {
> >       case IMX7D:
> > -     case IMX8MQ: /* FALLTHROUGH */
> > +             /* fall through */
> > +     case IMX8MQ:
> >               reset_control_assert(imx6_pcie->pciephy_reset);
> >               reset_control_assert(imx6_pcie->apps_reset);
> >               break;
>
> IMO this use of "fall through" is superfluous and unusual in the Linux
> source.
>
> A "fall through" comment would be useful if the IMX7D case had
> executable code but no "break".  Then the comment shows that the
> intent was to execute *both* the IMX7D code and the IMX8MQ code and
> the lack of a "break" was intentional.
>
> In this case, the intent is to treat IMX7D and IMX8MQ the same, and
> there's no executable code specifically for the IMX7D.  I think it's
> easiest to read that when the list of identical cases is all together
> without the comment in the middle, i.e., as
>
> >       case IMX7D:
> >       case IMX8MQ:
> >               reset_control_assert(imx6_pcie->pciephy_reset);
>
> rather than this:
>
> >       case IMX7D:
> >               /* fall through */
> >       case IMX8MQ:
> >               reset_control_assert(imx6_pcie->pciephy_reset);

OK, understood, will remove in next version.

Thanks,
Andrey Smirnov