[PATCH 0/2] DRA72/DRA74: Add 2 lane support

[PATCH 0/2] DRA72/DRA74: Add 2 lane support

[Kishon Vijay Abraham I]

Add driver modifications in pci-dra7xx to get x2 mode working in
DRA72 and DRA74. Certain modifications is needed in PHY driver also
which will be sent as a separate series.

Certain board modifications has to be done in order to test
x2 mode in dra72-evm.

These patches were created on pci next.

Changes from RFC:
*) .b1co_mode_sel_mask is now set with the correct value.
*) cleanup the patch

Kishon Vijay Abraham I (2):
  pci: host: pci-dra7xx: use "num-lanes" property to find phy count
  pci: host: pci-dra7xx: Enable x2 mode support

 Documentation/devicetree/bindings/pci/ti-pci.txt |    8 +-
 drivers/pci/host/pci-dra7xx.c                    |  104 +++++++++++++++++++---
 2 files changed, 97 insertions(+), 15 deletions(-)

-- 
1.7.9.5

--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

[PATCH 1/2] pci: host: pci-dra7xx: use "num-lanes" property to find phy count

[Kishon Vijay Abraham I]

use "num-lanes" property to find phy count instead of the number
phy-names property.

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
---
 drivers/pci/host/pci-dra7xx.c |   23 +++++++++++------------
 1 file changed, 11 insertions(+), 12 deletions(-)

diff --git a/drivers/pci/host/pci-dra7xx.c b/drivers/pci/host/pci-dra7xx.c
index 5963adc..05bbeee 100644
--- a/drivers/pci/host/pci-dra7xx.c
+++ b/drivers/pci/host/pci-dra7xx.c
@@ -70,7 +70,7 @@
 struct dra7xx_pcie {
 	void __iomem		*base;
 	struct phy		**phy;
-	int			phy_count;
+	int			lanes;
 	struct device		*dev;
 	struct pcie_port	pp;
 };
@@ -364,7 +364,7 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 	int ret;
 	int irq;
 	int i;
-	int phy_count;
+	u32 lanes;
 	struct phy **phy;
 	void __iomem *base;
 	struct resource *res;
@@ -402,17 +402,16 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 	if (!base)
 		return -ENOMEM;
 
-	phy_count = of_property_count_strings(np, "phy-names");
-	if (phy_count < 0) {
-		dev_err(dev, "unable to find the strings\n");
-		return phy_count;
+	if (of_property_read_u32(np, "num-lanes", &lanes)) {
+		dev_err(dev, "Failed to parse the number of lanes\n");
+		return -EINVAL;
 	}
 
-	phy = devm_kzalloc(dev, sizeof(*phy) * phy_count, GFP_KERNEL);
+	phy = devm_kzalloc(dev, sizeof(*phy) * lanes, GFP_KERNEL);
 	if (!phy)
 		return -ENOMEM;
 
-	for (i = 0; i < phy_count; i++) {
+	for (i = 0; i < lanes; i++) {
 		snprintf(name, sizeof(name), "pcie-phy%d", i);
 		phy[i] = devm_phy_get(dev, name);
 		if (IS_ERR(phy[i]))
@@ -432,7 +431,7 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 	dra7xx->base = base;
 	dra7xx->phy = phy;
 	dra7xx->dev = dev;
-	dra7xx->phy_count = phy_count;
+	dra7xx->lanes = lanes;
 
 	pm_runtime_enable(dev);
 	ret = pm_runtime_get_sync(dev);
@@ -489,7 +488,7 @@ static int __exit dra7xx_pcie_remove(struct platform_device *pdev)
 	struct dra7xx_pcie *dra7xx = platform_get_drvdata(pdev);
 	struct pcie_port *pp = &dra7xx->pp;
 	struct device *dev = &pdev->dev;
-	int count = dra7xx->phy_count;
+	int count = dra7xx->lanes;
 
 	if (pp->irq_domain)
 		irq_domain_remove(pp->irq_domain);
@@ -535,7 +534,7 @@ static int dra7xx_pcie_resume(struct device *dev)
 static int dra7xx_pcie_suspend_noirq(struct device *dev)
 {
 	struct dra7xx_pcie *dra7xx = dev_get_drvdata(dev);
-	int count = dra7xx->phy_count;
+	int count = dra7xx->lanes;
 
 	while (count--) {
 		phy_power_off(dra7xx->phy[count]);
@@ -548,7 +547,7 @@ static int dra7xx_pcie_suspend_noirq(struct device *dev)
 static int dra7xx_pcie_resume_noirq(struct device *dev)
 {
 	struct dra7xx_pcie *dra7xx = dev_get_drvdata(dev);
-	int phy_count = dra7xx->phy_count;
+	int phy_count = dra7xx->lanes;
 	int ret;
 	int i;
 
-- 
1.7.9.5

[PATCH 2/2] pci: host: pci-dra7xx: Enable x2 mode support

[Kishon Vijay Abraham I]

Perform syscon configurations to get x2 mode to working in DRA74x and
DRA72x. Also add a new compatible string to dfferentiate
DRA72x and DRA74x, since b1c0 mask is different for both these platforms.

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
---
 Documentation/devicetree/bindings/pci/ti-pci.txt |    8 ++-
 drivers/pci/host/pci-dra7xx.c                    |   81 +++++++++++++++++++++-
 2 files changed, 86 insertions(+), 3 deletions(-)

diff --git a/Documentation/devicetree/bindings/pci/ti-pci.txt b/Documentation/devicetree/bindings/pci/ti-pci.txt
index 60e2516..0b10e84 100644
--- a/Documentation/devicetree/bindings/pci/ti-pci.txt
+++ b/Documentation/devicetree/bindings/pci/ti-pci.txt
@@ -1,7 +1,9 @@
 TI PCI Controllers
 
 PCIe Designware Controller
- - compatible: Should be "ti,dra7-pcie""
+ - compatible: "ti,dra7-pcie" is deprecated
+	       Should be "ti,dra746-pcie" for DRA74x
+	       Should be "ti,dra726-pcie" for DRA72x
  - reg : Two register ranges as listed in the reg-names property
  - reg-names : The first entry must be "ti-conf" for the TI specific registers
 	       The second entry must be "rc-dbics" for the designware pcie
@@ -14,6 +16,10 @@ PCIe Designware Controller
 	       where <X> is the instance number of the pcie from the HW spec.
  - interrupts : Two interrupt entries must be specified. The first one is for
 		main interrupt line and the second for MSI interrupt line.
+ - syscon-lane-conf : phandle/offset pair. Phandle to the system control module and the
+   register offset to specify 1 lane or 2 lane.
+ - syscon-lane-sel : phandle/offset pair. Phandle to the system control module and the
+   register offset to specify lane selection.
  - #address-cells,
    #size-cells,
    #interrupt-cells,
diff --git a/drivers/pci/host/pci-dra7xx.c b/drivers/pci/host/pci-dra7xx.c
index 05bbeee..dac216f 100644
--- a/drivers/pci/host/pci-dra7xx.c
+++ b/drivers/pci/host/pci-dra7xx.c
@@ -22,9 +22,11 @@
 #include <linux/phy/phy.h>
 #include <linux/platform_device.h>
 #include <linux/pm_runtime.h>
+#include <linux/of_platform.h>
 #include <linux/resource.h>
 #include <linux/types.h>
-#include <linux/platform_data/pci-dra7xx.h>
+#include <linux/mfd/syscon.h>
+#include <linux/regmap.h>
 
 #include <linux/platform_data/pci-dra7xx.h>
 
@@ -67,14 +69,22 @@
 #define	LINK_UP						BIT(16)
 #define	DRA7XX_CPU_TO_BUS_ADDR				0x0FFFFFFF
 
+#define PCIE_1LANE_2LANE_SELECTION			BIT(13)
+#define PCIE_B1C0_MODE_SEL				BIT(2)
+
 struct dra7xx_pcie {
 	void __iomem		*base;
+	u32			*b1c0_mask;
 	struct phy		**phy;
 	int			lanes;
 	struct device		*dev;
 	struct pcie_port	pp;
 };
 
+struct dra7xx_pcie_data {
+	u32	b1co_mode_sel_mask;
+};
+
 #define to_dra7xx_pcie(x)	container_of((x), struct dra7xx_pcie, pp)
 
 static inline u32 dra7xx_pcie_readl(struct dra7xx_pcie *pcie, u32 offset)
@@ -358,6 +368,57 @@ static int dra7xx_pcie_reset(struct platform_device *pdev)
 	return 0;
 }
 
+static const struct of_device_id of_dra7xx_pcie_match[];
+
+static int dra7xx_pcie_configure_two_lane(struct device *dev)
+{
+	struct device_node *np = dev->of_node;
+	struct regmap *pcie_syscon;
+	unsigned int pcie_reg;
+	struct dra7xx_pcie_data *data;
+	const struct of_device_id *match;
+
+	match = of_match_device(of_dra7xx_pcie_match, dev);
+	if (!match)
+		return -EINVAL;
+
+	data = (struct dra7xx_pcie_data *)match->data;
+	if (!data) {
+		dev_err(dev, "no b1c0 mask data\n");
+		return -EINVAL;
+	}
+
+	pcie_syscon = syscon_regmap_lookup_by_phandle(np, "syscon-lane-conf");
+	if (IS_ERR(pcie_syscon)) {
+		dev_err(dev, "unable to get syscon-lane-conf\n");
+		return -EINVAL;
+	}
+
+	if (of_property_read_u32_index(np, "syscon-lane-conf", 1, &pcie_reg)) {
+		dev_err(dev, "couldn't get lane configuration reg offset\n");
+		return -EINVAL;
+	}
+
+	regmap_update_bits(pcie_syscon, pcie_reg, PCIE_1LANE_2LANE_SELECTION,
+			   PCIE_1LANE_2LANE_SELECTION);
+
+	pcie_syscon = syscon_regmap_lookup_by_phandle(np, "syscon-lane-sel");
+	if (IS_ERR(pcie_syscon)) {
+		dev_err(dev, "unable to get syscon-lane-sel\n");
+		return -EINVAL;
+	}
+
+	if (of_property_read_u32_index(np, "syscon-lane-sel", 1, &pcie_reg)) {
+		dev_err(dev, "couldn't get lane selection reg offset\n");
+		return -EINVAL;
+	}
+
+	regmap_update_bits(pcie_syscon, pcie_reg, data->b1co_mode_sel_mask,
+			   PCIE_B1C0_MODE_SEL);
+
+	return 0;
+}
+
 static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 {
 	u32 reg;
@@ -428,6 +489,12 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 		}
 	}
 
+	if (lanes == 2) {
+		ret = dra7xx_pcie_configure_two_lane(dev);
+		if (ret < 0)
+			goto err_phy;
+	}
+
 	dra7xx->base = base;
 	dra7xx->phy = phy;
 	dra7xx->dev = dev;
@@ -581,8 +648,18 @@ static const struct dev_pm_ops dra7xx_pcie_pm_ops = {
 				      dra7xx_pcie_resume_noirq)
 };
 
+static const struct dra7xx_pcie_data dra746_pcie_data = {
+	.b1co_mode_sel_mask = BIT(2),
+};
+
+static const struct dra7xx_pcie_data dra726_pcie_data = {
+	.b1co_mode_sel_mask = GENMASK(3, 2),
+};
+
 static const struct of_device_id of_dra7xx_pcie_match[] = {
-	{ .compatible = "ti,dra7-pcie", },
+	{ .compatible = "ti,dra7-pcie", .data = &dra746_pcie_data },
+	{ .compatible = "ti,dra746-pcie", .data = &dra746_pcie_data },
+	{ .compatible = "ti,dra726-pcie", .data = &dra726_pcie_data },
 	{},
 };
 MODULE_DEVICE_TABLE(of, of_dra7xx_pcie_match);
-- 
1.7.9.5

Re: [PATCH 2/2] pci: host: pci-dra7xx: Enable x2 mode support

[Rob Herring]

On Wed, Jan 06, 2016 at 04:19:53PM +0530, Kishon Vijay Abraham I wrote:
> Perform syscon configurations to get x2 mode to working in DRA74x and
> DRA72x. Also add a new compatible string to dfferentiate
> DRA72x and DRA74x, since b1c0 mask is different for both these platforms.
> 
> Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
> ---
>  Documentation/devicetree/bindings/pci/ti-pci.txt |    8 ++-
>  drivers/pci/host/pci-dra7xx.c                    |   81 +++++++++++++++++++++-
>  2 files changed, 86 insertions(+), 3 deletions(-)
> 
> diff --git a/Documentation/devicetree/bindings/pci/ti-pci.txt b/Documentation/devicetree/bindings/pci/ti-pci.txt
> index 60e2516..0b10e84 100644
> --- a/Documentation/devicetree/bindings/pci/ti-pci.txt
> +++ b/Documentation/devicetree/bindings/pci/ti-pci.txt
> @@ -1,7 +1,9 @@
>  TI PCI Controllers
>  
>  PCIe Designware Controller
> - - compatible: Should be "ti,dra7-pcie""
> + - compatible: "ti,dra7-pcie" is deprecated
> +	       Should be "ti,dra746-pcie" for DRA74x
> +	       Should be "ti,dra726-pcie" for DRA72x
>   - reg : Two register ranges as listed in the reg-names property
>   - reg-names : The first entry must be "ti-conf" for the TI specific registers
>  	       The second entry must be "rc-dbics" for the designware pcie
> @@ -14,6 +16,10 @@ PCIe Designware Controller
>  	       where <X> is the instance number of the pcie from the HW spec.
>   - interrupts : Two interrupt entries must be specified. The first one is for
>  		main interrupt line and the second for MSI interrupt line.
> + - syscon-lane-conf : phandle/offset pair. Phandle to the system control module and the
> +   register offset to specify 1 lane or 2 lane.
> + - syscon-lane-sel : phandle/offset pair. Phandle to the system control module and the
> +   register offset to specify lane selection.

These should have a ti prefix.

>   - #address-cells,
>     #size-cells,
>     #interrupt-cells,

Re: [PATCH 2/2] pci: host: pci-dra7xx: Enable x2 mode support

[Kishon Vijay Abraham I]

Hi,

On Wednesday 06 January 2016 07:43 PM, Rob Herring wrote:
> On Wed, Jan 06, 2016 at 04:19:53PM +0530, Kishon Vijay Abraham I wrote:
>> Perform syscon configurations to get x2 mode to working in DRA74x and
>> DRA72x. Also add a new compatible string to dfferentiate
>> DRA72x and DRA74x, since b1c0 mask is different for both these platforms.
>>
>> Signed-off-by: Kishon Vijay Abraham I <kishon-l0cyMroinI0@public.gmane.org>
>> ---
>>  Documentation/devicetree/bindings/pci/ti-pci.txt |    8 ++-
>>  drivers/pci/host/pci-dra7xx.c                    |   81 +++++++++++++++++++++-
>>  2 files changed, 86 insertions(+), 3 deletions(-)
>>
>> diff --git a/Documentation/devicetree/bindings/pci/ti-pci.txt b/Documentation/devicetree/bindings/pci/ti-pci.txt
>> index 60e2516..0b10e84 100644
>> --- a/Documentation/devicetree/bindings/pci/ti-pci.txt
>> +++ b/Documentation/devicetree/bindings/pci/ti-pci.txt
>> @@ -1,7 +1,9 @@
>>  TI PCI Controllers
>>  
>>  PCIe Designware Controller
>> - - compatible: Should be "ti,dra7-pcie""
>> + - compatible: "ti,dra7-pcie" is deprecated
>> +	       Should be "ti,dra746-pcie" for DRA74x
>> +	       Should be "ti,dra726-pcie" for DRA72x
>>   - reg : Two register ranges as listed in the reg-names property
>>   - reg-names : The first entry must be "ti-conf" for the TI specific registers
>>  	       The second entry must be "rc-dbics" for the designware pcie
>> @@ -14,6 +16,10 @@ PCIe Designware Controller
>>  	       where <X> is the instance number of the pcie from the HW spec.
>>   - interrupts : Two interrupt entries must be specified. The first one is for
>>  		main interrupt line and the second for MSI interrupt line.
>> + - syscon-lane-conf : phandle/offset pair. Phandle to the system control module and the
>> +   register offset to specify 1 lane or 2 lane.
>> + - syscon-lane-sel : phandle/offset pair. Phandle to the system control module and the
>> +   register offset to specify lane selection.
> 
> These should have a ti prefix.

Okay. Will fix that and post a new version.

Thanks
Kishon
> 
>>   - #address-cells,
>>     #size-cells,
>>     #interrupt-cells,
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 0/2] DRA72/DRA74: Add 2 lane support

[Bjorn Helgaas]

[+cc Richard]

Hi Kishon,

On Wed, Jan 06, 2016 at 04:19:51PM +0530, Kishon Vijay Abraham I wrote:
> Add driver modifications in pci-dra7xx to get x2 mode working in
> DRA72 and DRA74. Certain modifications is needed in PHY driver also
> which will be sent as a separate series.
> 
> Certain board modifications has to be done in order to test
> x2 mode in dra72-evm.
> 
> These patches were created on pci next.
> 
> Changes from RFC:
> *) .b1co_mode_sel_mask is now set with the correct value.
> *) cleanup the patch
> 
> Kishon Vijay Abraham I (2):
>   pci: host: pci-dra7xx: use "num-lanes" property to find phy count
>   pci: host: pci-dra7xx: Enable x2 mode support
> 
>  Documentation/devicetree/bindings/pci/ti-pci.txt |    8 +-
>  drivers/pci/host/pci-dra7xx.c                    |  104 +++++++++++++++++++---
>  2 files changed, 97 insertions(+), 15 deletions(-)

Apparently dra7xx in mainline doesn't work [1].  Until that's
resolved, I'm going to ignore dra7xx patches.

Bjorn

[1] http://lkml.kernel.org/r/20160106214518.GA6106-bi+AKbBUZKY6gyzm1THtWbp2dZbC/Bob@public.gmane.org
--
To unsubscribe from this list: send the line "unsubscribe devicetree" in
the body of a message to majordomo-u79uwXL29TY76Z2rM5mHXA@public.gmane.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: [PATCH 0/2] DRA72/DRA74: Add 2 lane support

[Tony Lindgren]

* Bjorn Helgaas <helgaas@kernel.org> [160107 09:42]:
> [+cc Richard]
> 
> Hi Kishon,
> 
> On Wed, Jan 06, 2016 at 04:19:51PM +0530, Kishon Vijay Abraham I wrote:
> > Add driver modifications in pci-dra7xx to get x2 mode working in
> > DRA72 and DRA74. Certain modifications is needed in PHY driver also
> > which will be sent as a separate series.
> > 
> > Certain board modifications has to be done in order to test
> > x2 mode in dra72-evm.
> > 
> > These patches were created on pci next.
> > 
> > Changes from RFC:
> > *) .b1co_mode_sel_mask is now set with the correct value.
> > *) cleanup the patch
> > 
> > Kishon Vijay Abraham I (2):
> >   pci: host: pci-dra7xx: use "num-lanes" property to find phy count
> >   pci: host: pci-dra7xx: Enable x2 mode support
> > 
> >  Documentation/devicetree/bindings/pci/ti-pci.txt |    8 +-
> >  drivers/pci/host/pci-dra7xx.c                    |  104 +++++++++++++++++++---
> >  2 files changed, 97 insertions(+), 15 deletions(-)
> 
> Apparently dra7xx in mainline doesn't work [1].  Until that's
> resolved, I'm going to ignore dra7xx patches.

Agreed. Guys, please do the development _and_ testing on the mainline
kernel. We do have the mainline kernel usable for probably a few
tens of omap SoC variants. If you have issues using dra7 with mainline,
please fix those issues ASAP.

Ideally of course please also use the device you're working on to
send out the patches :)

Regards,

Tony

> [1] http://lkml.kernel.org/r/20160106214518.GA6106@localhost.localdomain

Re: [PATCH 0/2] DRA72/DRA74: Add 2 lane support

[Kishon Vijay Abraham I]

Hi Tony,

On Thursday 07 January 2016 11:43 PM, Tony Lindgren wrote:
> * Bjorn Helgaas <helgaas@kernel.org> [160107 09:42]:
>> [+cc Richard]
>>
>> Hi Kishon,
>>
>> On Wed, Jan 06, 2016 at 04:19:51PM +0530, Kishon Vijay Abraham I wrote:
>>> Add driver modifications in pci-dra7xx to get x2 mode working in
>>> DRA72 and DRA74. Certain modifications is needed in PHY driver also
>>> which will be sent as a separate series.
>>>
>>> Certain board modifications has to be done in order to test
>>> x2 mode in dra72-evm.
>>>
>>> These patches were created on pci next.
>>>
>>> Changes from RFC:
>>> *) .b1co_mode_sel_mask is now set with the correct value.
>>> *) cleanup the patch
>>>
>>> Kishon Vijay Abraham I (2):
>>>   pci: host: pci-dra7xx: use "num-lanes" property to find phy count
>>>   pci: host: pci-dra7xx: Enable x2 mode support
>>>
>>>  Documentation/devicetree/bindings/pci/ti-pci.txt |    8 +-
>>>  drivers/pci/host/pci-dra7xx.c                    |  104 +++++++++++++++++++---
>>>  2 files changed, 97 insertions(+), 15 deletions(-)
>>
>> Apparently dra7xx in mainline doesn't work [1].  Until that's
>> resolved, I'm going to ignore dra7xx patches.
> 
> Agreed. Guys, please do the development _and_ testing on the mainline
> kernel. We do have the mainline kernel usable for probably a few
> tens of omap SoC variants. If you have issues using dra7 with mainline,
> please fix those issues ASAP.

Here Bjorn meant pci-dra7xx is broken in mainline and not the dra7xx boot
as such. (which is due to the absence of reset controller driver).
All the patches to pci-dra7xx were tested on mainline kernel (having
out-of-tree reset patches since the reset controller driver is not merged).
> 
> Ideally of course please also use the device you're working on to
> send out the patches :)

dra7x is the only device that use this pci driver and all the patches to
pci-dra7xx have been tested on both dra7 and dra72 boards.

Thanks
Kishon