[PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[Kishon Vijay Abraham I]

This series contains
        *) a patch to cleanup dra7xx probe
        *) a patch to force DRA7xx controller to work in GEN1 mode
        *) a patch to fix dra7xx driver so that even if CONFIG_PCI_MSI
           is enabled, dra7xx driver can service legacy interrupts
           (provided the device support only legacy interrupts).

Changes from v1:
*) rebased to 4.10-rc1 (and [1])
*) use existing dt property "max-link-speed" instead of creating a new
   TI specific dt property.

[1] -> https://lkml.org/lkml/2016/12/28/34

Kishon Vijay Abraham I (3):
  PCI: dra7xx: simplify the probe code
  PCI: dra7xx: Add support to force RC to work in GEN1 mode
  PCI: dra7xx: Enable MSI and legacy interrupts simultaneously

 drivers/pci/controller/pci-dra7xx.c |   81 ++++++++++++++++++++---------------
 1 file changed, 47 insertions(+), 34 deletions(-)

-- 
1.7.9.5

[PATCH v2 1/3] PCI: dra7xx: simplify the probe code

[Kishon Vijay Abraham I]

No functional change. Use the new devm_gpiod_get_optional() to
simplify the probe code.

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
Signed-off-by: Sekhar Nori <nsekhar@ti.com>
---
 drivers/pci/controller/pci-dra7xx.c |   21 +++++----------------
 1 file changed, 5 insertions(+), 16 deletions(-)

diff --git a/drivers/pci/controller/pci-dra7xx.c b/drivers/pci/controller/pci-dra7xx.c
index 9595fad..f6d0c63 100644
--- a/drivers/pci/controller/pci-dra7xx.c
+++ b/drivers/pci/controller/pci-dra7xx.c
@@ -320,9 +320,7 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 	struct device *dev = &pdev->dev;
 	struct device_node *np = dev->of_node;
 	char name[10];
-	int gpio_sel;
-	enum of_gpio_flags flags;
-	unsigned long gpio_flags;
+	struct gpio_desc *reset;
 
 	dra7xx = devm_kzalloc(dev, sizeof(*dra7xx), GFP_KERNEL);
 	if (!dra7xx)
@@ -388,19 +386,10 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 		goto err_get_sync;
 	}
 
-	gpio_sel = of_get_gpio_flags(dev->of_node, 0, &flags);
-	if (gpio_is_valid(gpio_sel)) {
-		gpio_flags = (flags & OF_GPIO_ACTIVE_LOW) ?
-				GPIOF_OUT_INIT_LOW : GPIOF_OUT_INIT_HIGH;
-		ret = devm_gpio_request_one(dev, gpio_sel, gpio_flags,
-					    "pcie_reset");
-		if (ret) {
-			dev_err(dev, "gpio%d request failed, ret %d\n",
-				gpio_sel, ret);
-			goto err_gpio;
-		}
-	} else if (gpio_sel == -EPROBE_DEFER) {
-		ret = -EPROBE_DEFER;
+	reset = devm_gpiod_get_optional(dev, NULL, GPIOD_OUT_HIGH);
+	if (IS_ERR(reset)) {
+		ret = PTR_ERR(reset);
+		dev_err(&pdev->dev, "gpio request failed, ret %d\n", ret);
 		goto err_gpio;
 	}
 
-- 
1.7.9.5

[PATCH v2 2/3] PCI: dra7xx: Add support to force RC to work in GEN1 mode

[Kishon Vijay Abraham I]

PCIe in AM57x/DRA7x devices is by default
configured to work in GEN2 mode.  However there
may be situations when working in GEN1 mode is
desired. One example is limitation i925 (PCIe GEN2
mode not supported at junction temperatures < 0C).

Add support to force Root Complex to work in GEN1
mode if so desired, but don't force GEN1 mode on
any board just yet.

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
---
 drivers/pci/controller/pci-dra7xx.c |   29 +++++++++++++++++++++++++++++
 1 file changed, 29 insertions(+)

diff --git a/drivers/pci/controller/pci-dra7xx.c b/drivers/pci/controller/pci-dra7xx.c
index f6d0c63..587b18c 100644
--- a/drivers/pci/controller/pci-dra7xx.c
+++ b/drivers/pci/controller/pci-dra7xx.c
@@ -17,6 +17,7 @@
 #include <linux/kernel.h>
 #include <linux/init.h>
 #include <linux/of_gpio.h>
+#include <linux/of_pci.h>
 #include <linux/pci.h>
 #include <linux/phy/phy.h>
 #include <linux/platform_device.h>
@@ -63,11 +64,14 @@
 #define	LINK_UP						BIT(16)
 #define	DRA7XX_CPU_TO_BUS_ADDR				0x0FFFFFFF
 
+#define EXP_CAP_ID_OFFSET				0x70
+
 struct dra7xx_pcie {
 	struct pcie_port	pp;
 	void __iomem		*base;		/* DT ti_conf */
 	int			phy_count;	/* DT phy-names count */
 	struct phy		**phy;
+	int			link_gen;
 };
 
 #define to_dra7xx_pcie(x)	container_of((x), struct dra7xx_pcie, pp)
@@ -96,12 +100,33 @@ static int dra7xx_pcie_establish_link(struct dra7xx_pcie *dra7xx)
 	struct pcie_port *pp = &dra7xx->pp;
 	struct device *dev = pp->dev;
 	u32 reg;
+	u32 exp_cap_off = EXP_CAP_ID_OFFSET;
 
 	if (dw_pcie_link_up(pp)) {
 		dev_err(dev, "link is already up\n");
 		return 0;
 	}
 
+	if (dra7xx->link_gen == 1) {
+		dw_pcie_cfg_read(pp->dbi_base + exp_cap_off + PCI_EXP_LNKCAP,
+				 4, &reg);
+		if ((reg & PCI_EXP_LNKCAP_SLS) != PCI_EXP_LNKCAP_SLS_2_5GB) {
+			reg &= ~((u32)PCI_EXP_LNKCAP_SLS);
+			reg |= PCI_EXP_LNKCAP_SLS_2_5GB;
+			dw_pcie_cfg_write(pp->dbi_base + exp_cap_off +
+					  PCI_EXP_LNKCAP, 4, reg);
+		}
+
+		dw_pcie_cfg_read(pp->dbi_base + exp_cap_off + PCI_EXP_LNKCTL2,
+				 2, &reg);
+		if ((reg & PCI_EXP_LNKCAP_SLS) != PCI_EXP_LNKCAP_SLS_2_5GB) {
+			reg &= ~((u32)PCI_EXP_LNKCAP_SLS);
+			reg |= PCI_EXP_LNKCAP_SLS_2_5GB;
+			dw_pcie_cfg_write(pp->dbi_base + exp_cap_off +
+					  PCI_EXP_LNKCTL2, 2, reg);
+		}
+	}
+
 	reg = dra7xx_pcie_readl(dra7xx, PCIECTRL_DRA7XX_CONF_DEVICE_CMD);
 	reg |= LTSSM_EN;
 	dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_DEVICE_CMD, reg);
@@ -397,6 +422,10 @@ static int __init dra7xx_pcie_probe(struct platform_device *pdev)
 	reg &= ~LTSSM_EN;
 	dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_DEVICE_CMD, reg);
 
+	dra7xx->link_gen = of_pci_get_max_link_speed(np);
+	if (dra7xx->link_gen < 0 || dra7xx->link_gen > 2)
+		dra7xx->link_gen = 2;
+
 	ret = dra7xx_add_pcie_port(dra7xx, pdev);
 	if (ret < 0)
 		goto err_gpio;
-- 
1.7.9.5

[PATCH v2 3/3] PCI: dra7xx: Enable MSI and legacy interrupts simultaneously

[Kishon Vijay Abraham I]

dra7xx driver had a bug in that if CONFIG_PCI_MSI config is enabled,
it doesn't support legacy interrupt. Fix it here so that both MSI and
legacy interrupts can be enabled simultaneously and one of them will
be used based on the connected device.

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
---
 drivers/pci/controller/pci-dra7xx.c |   31 +++++++++++++------------------
 1 file changed, 13 insertions(+), 18 deletions(-)

diff --git a/drivers/pci/controller/pci-dra7xx.c b/drivers/pci/controller/pci-dra7xx.c
index 587b18c..5ee14b7 100644
--- a/drivers/pci/controller/pci-dra7xx.c
+++ b/drivers/pci/controller/pci-dra7xx.c
@@ -72,6 +72,7 @@ struct dra7xx_pcie {
 	int			phy_count;	/* DT phy-names count */
 	struct phy		**phy;
 	int			link_gen;
+	struct irq_domain	*irq_domain;
 };
 
 #define to_dra7xx_pcie(x)	container_of((x), struct dra7xx_pcie, pp)
@@ -143,13 +144,8 @@ static void dra7xx_pcie_enable_interrupts(struct dra7xx_pcie *dra7xx)
 	dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQSTATUS_MSI,
 			   ~LEG_EP_INTERRUPTS & ~MSI);
 
-	if (IS_ENABLED(CONFIG_PCI_MSI))
-		dra7xx_pcie_writel(dra7xx,
-				   PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MSI, MSI);
-	else
-		dra7xx_pcie_writel(dra7xx,
-				   PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MSI,
-				   LEG_EP_INTERRUPTS);
+	dra7xx_pcie_writel(dra7xx, PCIECTRL_DRA7XX_CONF_IRQENABLE_SET_MSI,
+			   MSI | LEG_EP_INTERRUPTS);
 }
 
 static void dra7xx_pcie_host_init(struct pcie_port *pp)
@@ -164,8 +160,7 @@ static void dra7xx_pcie_host_init(struct pcie_port *pp)
 	dw_pcie_setup_rc(pp);
 
 	dra7xx_pcie_establish_link(dra7xx);
-	if (IS_ENABLED(CONFIG_PCI_MSI))
-		dw_pcie_msi_init(pp);
+	dw_pcie_msi_init(pp);
 	dra7xx_pcie_enable_interrupts(dra7xx);
 }
 
@@ -190,6 +185,7 @@ static int dra7xx_pcie_intx_map(struct irq_domain *domain, unsigned int irq,
 static int dra7xx_pcie_init_irq_domain(struct pcie_port *pp)
 {
 	struct device *dev = pp->dev;
+	struct dra7xx_pcie *dra7xx_pcie = to_dra7xx_pcie(pp);
 	struct device_node *node = dev->of_node;
 	struct device_node *pcie_intc_node =  of_get_next_child(node, NULL);
 
@@ -198,9 +194,9 @@ static int dra7xx_pcie_init_irq_domain(struct pcie_port *pp)
 		return -ENODEV;
 	}
 
-	pp->irq_domain = irq_domain_add_linear(pcie_intc_node, 4,
-					       &intx_domain_ops, pp);
-	if (!pp->irq_domain) {
+	dra7xx_pcie->irq_domain = irq_domain_add_linear(pcie_intc_node, 4,
+							&intx_domain_ops, pp);
+	if (!dra7xx_pcie->irq_domain) {
 		dev_err(dev, "Failed to get a INTx IRQ domain\n");
 		return -ENODEV;
 	}
@@ -224,7 +220,8 @@ static irqreturn_t dra7xx_pcie_msi_irq_handler(int irq, void *arg)
 	case INTB:
 	case INTC:
 	case INTD:
-		generic_handle_irq(irq_find_mapping(pp->irq_domain, ffs(reg)));
+		generic_handle_irq(irq_find_mapping(dra7xx->irq_domain,
+						    ffs(reg)));
 		break;
 	}
 
@@ -310,11 +307,9 @@ static int __init dra7xx_add_pcie_port(struct dra7xx_pcie *dra7xx,
 		return ret;
 	}
 
-	if (!IS_ENABLED(CONFIG_PCI_MSI)) {
-		ret = dra7xx_pcie_init_irq_domain(pp);
-		if (ret < 0)
-			return ret;
-	}
+	ret = dra7xx_pcie_init_irq_domain(pp);
+	if (ret < 0)
+		return ret;
 
 	res = platform_get_resource_byname(pdev, IORESOURCE_MEM, "rc_dbics");
 	pp->dbi_base = devm_ioremap(dev, res->start, resource_size(res));
-- 
1.7.9.5

Re: [PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[Bjorn Helgaas]

On Fri, Dec 30, 2016 at 03:26:11PM +0530, Kishon Vijay Abraham I wrote:
> This series contains
>         *) a patch to cleanup dra7xx probe
>         *) a patch to force DRA7xx controller to work in GEN1 mode
>         *) a patch to fix dra7xx driver so that even if CONFIG_PCI_MSI
>            is enabled, dra7xx driver can service legacy interrupts
>            (provided the device support only legacy interrupts).
> 
> Changes from v1:
> *) rebased to 4.10-rc1 (and [1])
> *) use existing dt property "max-link-speed" instead of creating a new
>    TI specific dt property.
> 
> [1] -> https://lkml.org/lkml/2016/12/28/34
> 
> Kishon Vijay Abraham I (3):
>   PCI: dra7xx: simplify the probe code
>   PCI: dra7xx: Add support to force RC to work in GEN1 mode
>   PCI: dra7xx: Enable MSI and legacy interrupts simultaneously
> 
>  drivers/pci/controller/pci-dra7xx.c |   81 ++++++++++++++++++++---------------
>  1 file changed, 47 insertions(+), 34 deletions(-)

Applied to pci/host-dra7xx for v4.11.  I haven't applied the directory
rename yet, so I'll fix that up as necessary.

Bjorn

Re: [PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[Kishon Vijay Abraham I]

Hi Bjorn,

On Thursday 12 January 2017 02:51 AM, Bjorn Helgaas wrote:
> On Fri, Dec 30, 2016 at 03:26:11PM +0530, Kishon Vijay Abraham I wrote:
>> This series contains
>>         *) a patch to cleanup dra7xx probe
>>         *) a patch to force DRA7xx controller to work in GEN1 mode
>>         *) a patch to fix dra7xx driver so that even if CONFIG_PCI_MSI
>>            is enabled, dra7xx driver can service legacy interrupts
>>            (provided the device support only legacy interrupts).
>>
>> Changes from v1:
>> *) rebased to 4.10-rc1 (and [1])
>> *) use existing dt property "max-link-speed" instead of creating a new
>>    TI specific dt property.
>>
>> [1] -> https://lkml.org/lkml/2016/12/28/34
>>
>> Kishon Vijay Abraham I (3):
>>   PCI: dra7xx: simplify the probe code
>>   PCI: dra7xx: Add support to force RC to work in GEN1 mode
>>   PCI: dra7xx: Enable MSI and legacy interrupts simultaneously
>>
>>  drivers/pci/controller/pci-dra7xx.c |   81 ++++++++++++++++++++---------------
>>  1 file changed, 47 insertions(+), 34 deletions(-)
> 
> Applied to pci/host-dra7xx for v4.11.  I haven't applied the directory
> rename yet, so I'll fix that up as necessary.

I'd have liked that to be merged as well. Is that because Arnd hasn't Acked
that one yet?

Thanks
Kishon

Re: [PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[Bjorn Helgaas]

On Thu, Jan 12, 2017 at 10:52:56AM +0530, Kishon Vijay Abraham I wrote:
> Hi Bjorn,
> 
> On Thursday 12 January 2017 02:51 AM, Bjorn Helgaas wrote:
> > On Fri, Dec 30, 2016 at 03:26:11PM +0530, Kishon Vijay Abraham I wrote:
> >> This series contains
> >>         *) a patch to cleanup dra7xx probe
> >>         *) a patch to force DRA7xx controller to work in GEN1 mode
> >>         *) a patch to fix dra7xx driver so that even if CONFIG_PCI_MSI
> >>            is enabled, dra7xx driver can service legacy interrupts
> >>            (provided the device support only legacy interrupts).
> >>
> >> Changes from v1:
> >> *) rebased to 4.10-rc1 (and [1])
> >> *) use existing dt property "max-link-speed" instead of creating a new
> >>    TI specific dt property.
> >>
> >> [1] -> https://lkml.org/lkml/2016/12/28/34
> >>
> >> Kishon Vijay Abraham I (3):
> >>   PCI: dra7xx: simplify the probe code
> >>   PCI: dra7xx: Add support to force RC to work in GEN1 mode
> >>   PCI: dra7xx: Enable MSI and legacy interrupts simultaneously
> >>
> >>  drivers/pci/controller/pci-dra7xx.c |   81 ++++++++++++++++++++---------------
> >>  1 file changed, 47 insertions(+), 34 deletions(-)
> > 
> > Applied to pci/host-dra7xx for v4.11.  I haven't applied the directory
> > rename yet, so I'll fix that up as necessary.
> 
> I'd have liked that to be merged as well. Is that because Arnd hasn't Acked
> that one yet?

Nope, I expect to do the rename in some form.  I'm just trying to
clear out the easy stuff first before moving on to the harder patches
in the queue.

Bjorn

Re: [PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[Kishon Vijay Abraham I]

Hi,

On Thursday 12 January 2017 06:45 PM, Bjorn Helgaas wrote:
> On Thu, Jan 12, 2017 at 10:52:56AM +0530, Kishon Vijay Abraham I wrote:
>> Hi Bjorn,
>>
>> On Thursday 12 January 2017 02:51 AM, Bjorn Helgaas wrote:
>>> On Fri, Dec 30, 2016 at 03:26:11PM +0530, Kishon Vijay Abraham I wrote:
>>>> This series contains
>>>>         *) a patch to cleanup dra7xx probe
>>>>         *) a patch to force DRA7xx controller to work in GEN1 mode
>>>>         *) a patch to fix dra7xx driver so that even if CONFIG_PCI_MSI
>>>>            is enabled, dra7xx driver can service legacy interrupts
>>>>            (provided the device support only legacy interrupts).
>>>>
>>>> Changes from v1:
>>>> *) rebased to 4.10-rc1 (and [1])
>>>> *) use existing dt property "max-link-speed" instead of creating a new
>>>>    TI specific dt property.
>>>>
>>>> [1] -> https://lkml.org/lkml/2016/12/28/34
>>>>
>>>> Kishon Vijay Abraham I (3):
>>>>   PCI: dra7xx: simplify the probe code
>>>>   PCI: dra7xx: Add support to force RC to work in GEN1 mode
>>>>   PCI: dra7xx: Enable MSI and legacy interrupts simultaneously
>>>>
>>>>  drivers/pci/controller/pci-dra7xx.c |   81 ++++++++++++++++++++---------------
>>>>  1 file changed, 47 insertions(+), 34 deletions(-)
>>>
>>> Applied to pci/host-dra7xx for v4.11.  I haven't applied the directory
>>> rename yet, so I'll fix that up as necessary.
>>
>> I'd have liked that to be merged as well. Is that because Arnd hasn't Acked
>> that one yet?
> 
> Nope, I expect to do the rename in some form.  I'm just trying to
> clear out the easy stuff first before moving on to the harder patches
> in the queue.

cool, thanks. Let me know if I have to resend the patches based on one of your
branches (i.e after you apply the easy stuff).

Thanks
Kishon

Re: [PATCH v2 0/3] PCI: designware/dra7xx: misc fixes and cleanups

[Kishon Vijay Abraham I]

Hi,

On Thursday 12 January 2017 06:53 PM, Kishon Vijay Abraham I wrote:
> Hi,
> 
> On Thursday 12 January 2017 06:45 PM, Bjorn Helgaas wrote:
>> On Thu, Jan 12, 2017 at 10:52:56AM +0530, Kishon Vijay Abraham I wrote:
>>> Hi Bjorn,
>>>
>>> On Thursday 12 January 2017 02:51 AM, Bjorn Helgaas wrote:
>>>> On Fri, Dec 30, 2016 at 03:26:11PM +0530, Kishon Vijay Abraham I wrote:
>>>>> This series contains
>>>>>         *) a patch to cleanup dra7xx probe
>>>>>         *) a patch to force DRA7xx controller to work in GEN1 mode
>>>>>         *) a patch to fix dra7xx driver so that even if CONFIG_PCI_MSI
>>>>>            is enabled, dra7xx driver can service legacy interrupts
>>>>>            (provided the device support only legacy interrupts).
>>>>>
>>>>> Changes from v1:
>>>>> *) rebased to 4.10-rc1 (and [1])
>>>>> *) use existing dt property "max-link-speed" instead of creating a new
>>>>>    TI specific dt property.
>>>>>
>>>>> [1] -> https://lkml.org/lkml/2016/12/28/34
>>>>>
>>>>> Kishon Vijay Abraham I (3):
>>>>>   PCI: dra7xx: simplify the probe code
>>>>>   PCI: dra7xx: Add support to force RC to work in GEN1 mode
>>>>>   PCI: dra7xx: Enable MSI and legacy interrupts simultaneously
>>>>>
>>>>>  drivers/pci/controller/pci-dra7xx.c |   81 ++++++++++++++++++++---------------
>>>>>  1 file changed, 47 insertions(+), 34 deletions(-)
>>>>
>>>> Applied to pci/host-dra7xx for v4.11.  I haven't applied the directory
>>>> rename yet, so I'll fix that up as necessary.
>>>
>>> I'd have liked that to be merged as well. Is that because Arnd hasn't Acked
>>> that one yet?
>>
>> Nope, I expect to do the rename in some form.  I'm just trying to
>> clear out the easy stuff first before moving on to the harder patches
>> in the queue.
> 
> cool, thanks. Let me know if I have to resend the patches based on one of your
> branches (i.e after you apply the easy stuff).

Btw, I had sent a new patch series [1] which had a couple of additional patches
(that was after creating drivers/pci/dwc/). Consider merging that too or I can
resend the missing patches.

Thanks
Kishon

[1] -> https://lkml.org/lkml/2017/1/11/238
> 
> Thanks
> Kishon
>