[PATCH v4 0/6] PCI: rcar: Failpath fixes

[PATCH v4 0/6] PCI: rcar: Failpath fixes

[Marek Vasut]

Multiple minor failpath fixes for the R-Car PCIe driver.

V4: Sync up the version numbers
    Rebase on top of Lorenzo's tree
    Add new patch fixing bug in the PHY code

Marek Vasut (6):
  PCI: rcar: Pull bus clock enable/disable from
    rcar_pcie_get_resources()
  PCI: rcar: Add missing irq_dispose_mapping() into failpath
  PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails
  PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()
  PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath
  PCI: rcar: Shut the PHY down in failpath

 drivers/pci/host/pcie-rcar.c | 82 +++++++++++++++++++++++++++++++++++---------
 1 file changed, 66 insertions(+), 16 deletions(-)

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org

-- 
2.16.2

[PATCH v4 1/6] PCI: rcar: Pull bus clock enable/disable from rcar_pcie_get_resources()

[Marek Vasut]

The rcar_pcie_get_resources() is another misnomer with a side effect.
The function does not only get resources, but also enables/disables bus
clock. This is forgotten in the probe() function though and if anything
in probe() fails after rcar_pcie_get_resources() is called, the bus
clock are never disabled.

This patch pulls the clock handling out of the rcar_pcie_get_resources()
and enables clock after all the resources were requested. Moreover, this
patch also always disables the clock in case of failure.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org
---
V2: No change
V3: No change
V4: Rebase on top of Lorenzo's pci/rcar, fix up the rebase breakage
    induced by patches therein
---
 drivers/pci/host/pcie-rcar.c | 33 ++++++++++++++++-----------------
 1 file changed, 16 insertions(+), 17 deletions(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 8d161563bce8..0879880f7975 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -946,32 +946,22 @@ static int rcar_pcie_get_resources(struct rcar_pcie *pcie)
 		dev_err(dev, "cannot get pcie bus clock\n");
 		return PTR_ERR(pcie->bus_clk);
 	}
-	err = clk_prepare_enable(pcie->bus_clk);
-	if (err)
-		return err;
 
 	i = irq_of_parse_and_map(dev->of_node, 0);
 	if (!i) {
 		dev_err(dev, "cannot get platform resources for msi interrupt\n");
-		err = -ENOENT;
-		goto err_map_reg;
+		return -ENOENT;
 	}
 	pcie->msi.irq1 = i;
 
 	i = irq_of_parse_and_map(dev->of_node, 1);
 	if (!i) {
 		dev_err(dev, "cannot get platform resources for msi interrupt\n");
-		err = -ENOENT;
-		goto err_map_reg;
+		return -ENOENT;
 	}
 	pcie->msi.irq2 = i;
 
 	return 0;
-
-err_map_reg:
-	clk_disable_unprepare(pcie->bus_clk);
-
-	return err;
 }
 
 static int rcar_pcie_inbound_ranges(struct rcar_pcie *pcie,
@@ -1115,22 +1105,28 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 		goto err_pm_put;
 	}
 
+	err = clk_prepare_enable(pcie->bus_clk);
+	if (err) {
+		dev_err(dev, "failed to enable bus clock: %d\n", err);
+		goto err_pm_put;
+	}
+
 	err = rcar_pcie_parse_map_dma_ranges(pcie, dev->of_node);
 	if (err)
-		goto err_pm_put;
+		goto err_clk_disable;
 
 	phy_init_fn = of_device_get_match_data(dev);
 	err = phy_init_fn(pcie);
 	if (err) {
 		dev_err(dev, "failed to init PCIe PHY\n");
-		goto err_pm_put;
+		goto err_clk_disable;
 	}
 
 	/* Failure to get a link might just be that no cards are inserted */
 	if (rcar_pcie_hw_init(pcie)) {
 		dev_info(dev, "PCIe link down\n");
 		err = -ENODEV;
-		goto err_pm_put;
+		goto err_clk_disable;
 	}
 
 	data = rcar_pci_read_reg(pcie, MACSR);
@@ -1142,16 +1138,19 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 			dev_err(dev,
 				"failed to enable MSI support: %d\n",
 				err);
-			goto err_pm_put;
+			goto err_clk_disable;
 		}
 	}
 
 	err = rcar_pcie_enable(pcie);
 	if (err)
-		goto err_pm_put;
+		goto err_clk_disable;
 
 	return 0;
 
+err_clk_disable:
+	clk_disable_unprepare(pcie->bus_clk);
+
 err_pm_put:
 	pm_runtime_put(dev);
 
-- 
2.16.2

[PATCH v4 2/6] PCI: rcar: Add missing irq_dispose_mapping() into failpath

[Marek Vasut]

The rcar_pcie_get_resources() is another misnomer with a side effect.
The function does not only get resources, but also maps MSI IRQs via
irq_of_parse_and_map(). In case anything fails afterward, the IRQ
mapping must be disposed through irq_dispose_mapping() which is not
done.

This patch handles irq_of_parse_and_map() failures in by disposing
of the mapping in rcar_pcie_get_resources() as well as in probe.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org
---
V2: No change
V3: No change
V4: Rebase on top of Lorenzo's pci/rcar, fix up the rebase breakage
    induced by patches therein
---
 drivers/pci/host/pcie-rcar.c | 17 ++++++++++++++---
 1 file changed, 14 insertions(+), 3 deletions(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 0879880f7975..6bb1cdec26be 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -950,18 +950,25 @@ static int rcar_pcie_get_resources(struct rcar_pcie *pcie)
 	i = irq_of_parse_and_map(dev->of_node, 0);
 	if (!i) {
 		dev_err(dev, "cannot get platform resources for msi interrupt\n");
-		return -ENOENT;
+		err = -ENOENT;
+		goto err_irq1;
 	}
 	pcie->msi.irq1 = i;
 
 	i = irq_of_parse_and_map(dev->of_node, 1);
 	if (!i) {
 		dev_err(dev, "cannot get platform resources for msi interrupt\n");
-		return -ENOENT;
+		err = -ENOENT;
+		goto err_irq2;
 	}
 	pcie->msi.irq2 = i;
 
 	return 0;
+
+err_irq2:
+	irq_dispose_mapping(pcie->msi.irq1);
+err_irq1:
+	return err;
 }
 
 static int rcar_pcie_inbound_ranges(struct rcar_pcie *pcie,
@@ -1108,7 +1115,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 	err = clk_prepare_enable(pcie->bus_clk);
 	if (err) {
 		dev_err(dev, "failed to enable bus clock: %d\n", err);
-		goto err_pm_put;
+		goto err_unmap_msi_irqs;
 	}
 
 	err = rcar_pcie_parse_map_dma_ranges(pcie, dev->of_node);
@@ -1151,6 +1158,10 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 err_clk_disable:
 	clk_disable_unprepare(pcie->bus_clk);
 
+err_unmap_msi_irqs:
+	irq_dispose_mapping(pcie->msi.irq2);
+	irq_dispose_mapping(pcie->msi.irq1);
+
 err_pm_put:
 	pm_runtime_put(dev);
 
-- 
2.16.2

[PATCH v4 3/6] PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails

[Marek Vasut]

If the rcar_pcie_enable() fails and MSIs are enabled, the setup done in
rcar_pcie_enable_msi() is never undone. Add a function to tear down the
MSI setup by disabling the MSI handling in the PCIe block, deallocating
the pages requested for the MSIs and zapping the IRQ mapping.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org
---
V2: No change
V3: No change
V4: Rebase on top of Lorenzo's pci/rcar, fix up the rebase breakage
    induced by patches therein
---
 drivers/pci/host/pcie-rcar.c | 28 +++++++++++++++++++++++++++-
 1 file changed, 27 insertions(+), 1 deletion(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 6bb1cdec26be..d6a7e8482dcf 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -923,6 +923,28 @@ static int rcar_pcie_enable_msi(struct rcar_pcie *pcie)
 	return err;
 }
 
+static void rcar_pcie_teardown_msi(struct rcar_pcie *pcie)
+{
+	struct rcar_msi *msi = &pcie->msi;
+	int irq, i;
+
+	/* Disable all MSI interrupts */
+	rcar_pci_write_reg(pcie, 0, PCIEMSIIER);
+
+	/* Disable address decoding of the MSI interrupt, MSIFE */
+	rcar_pci_write_reg(pcie, 0, PCIEMSIALR);
+
+	free_pages(msi->pages, 0);
+
+	for (i = 0; i < INT_PCI_MSI_NR; i++) {
+		irq = irq_find_mapping(msi->domain, i);
+		if (irq > 0)
+			irq_dispose_mapping(irq);
+	}
+
+	irq_domain_remove(msi->domain);
+}
+
 static int rcar_pcie_get_resources(struct rcar_pcie *pcie)
 {
 	struct device *dev = pcie->dev;
@@ -1151,10 +1173,14 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 
 	err = rcar_pcie_enable(pcie);
 	if (err)
-		goto err_clk_disable;
+		goto err_msi_teardown;
 
 	return 0;
 
+err_msi_teardown:
+	if (IS_ENABLED(CONFIG_PCI_MSI))
+		rcar_pcie_teardown_msi(pcie);
+
 err_clk_disable:
 	clk_disable_unprepare(pcie->bus_clk);
 
-- 
2.16.2

[PATCH v4 4/6] PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()

[Marek Vasut]

The data link active signal usually takes ~20 uSec to be asserted,
poll the bit more often to avoid useless delays in this function.
Use udelay() instead of usleep() for such a small delay as suggested
by the timer documentation and because this will be used in atomic
context later on when the suspend/resume patches land.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org
---
V2: s/content/context in commit message
V3: Add cpu_relax()
V4: Rebase on top of Lorenzo's pci/rcar, fix up the rebase breakage
    induced by patches therein
---
 drivers/pci/host/pcie-rcar.c | 5 +++--
 1 file changed, 3 insertions(+), 2 deletions(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index d6a7e8482dcf..4843a4dc6059 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -546,13 +546,14 @@ static int rcar_pcie_wait_for_phyrdy(struct rcar_pcie *pcie)
 
 static int rcar_pcie_wait_for_dl(struct rcar_pcie *pcie)
 {
-	unsigned int timeout = 10;
+	unsigned int timeout = 10000;
 
 	while (timeout--) {
 		if ((rcar_pci_read_reg(pcie, PCIETSTR) & DATA_LINK_ACTIVE))
 			return 0;
 
-		msleep(5);
+		udelay(5);
+		cpu_relax();
 	}
 
 	return -ETIMEDOUT;
-- 
2.16.2

[PATCH v4 5/6] PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath

[Marek Vasut]

The rcar_pcie_enable_msi() creates IRQ mappings using irq_create_mapping()
before requesting the IRQs using devm_request_irq(). If devm_request_irq()
fails for some reason, rcar_pcie_enable_msi() does not remove the mapping.

Pull out the code for disposing IRQ mappings from rcar_pcie_teardown_msi()
into a separate function and call it from both rcar_pcie_teardown_msi()
and rcar_pcie_enable_msi() failpath to remove the mappings correctly.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org
---
V2: No change
V3: No change
V4: Rebase on top of Lorenzo's pci/rcar, fix up the rebase breakage
    induced by patches therein
---
 drivers/pci/host/pcie-rcar.c | 25 ++++++++++++++++---------
 1 file changed, 16 insertions(+), 9 deletions(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 4843a4dc6059..636c3c5095d2 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -866,6 +866,20 @@ static const struct irq_domain_ops msi_domain_ops = {
 	.map = rcar_msi_map,
 };
 
+static void rcar_pcie_unmap_msi(struct rcar_pcie *pcie)
+{
+	struct rcar_msi *msi = &pcie->msi;
+	int i, irq;
+
+	for (i = 0; i < INT_PCI_MSI_NR; i++) {
+		irq = irq_find_mapping(msi->domain, i);
+		if (irq > 0)
+			irq_dispose_mapping(irq);
+	}
+
+	irq_domain_remove(msi->domain);
+}
+
 static int rcar_pcie_enable_msi(struct rcar_pcie *pcie)
 {
 	struct device *dev = pcie->dev;
@@ -920,14 +934,13 @@ static int rcar_pcie_enable_msi(struct rcar_pcie *pcie)
 	return 0;
 
 err:
-	irq_domain_remove(msi->domain);
+	rcar_pcie_unmap_msi(pcie);
 	return err;
 }
 
 static void rcar_pcie_teardown_msi(struct rcar_pcie *pcie)
 {
 	struct rcar_msi *msi = &pcie->msi;
-	int irq, i;
 
 	/* Disable all MSI interrupts */
 	rcar_pci_write_reg(pcie, 0, PCIEMSIIER);
@@ -937,13 +950,7 @@ static void rcar_pcie_teardown_msi(struct rcar_pcie *pcie)
 
 	free_pages(msi->pages, 0);
 
-	for (i = 0; i < INT_PCI_MSI_NR; i++) {
-		irq = irq_find_mapping(msi->domain, i);
-		if (irq > 0)
-			irq_dispose_mapping(irq);
-	}
-
-	irq_domain_remove(msi->domain);
+	rcar_pcie_unmap_msi(pcie);
 }
 
 static int rcar_pcie_get_resources(struct rcar_pcie *pcie)
-- 
2.16.2

[PATCH v4 6/6] PCI: rcar: Shut the PHY down in failpath

[Marek Vasut]

If anything fails past phy_init_fn() and the system is a Gen3 with
a PHY, the PHY will be left on and inited. This is caused by the
phy_init_fn, which is in fact a pointer to rcar_pcie_phy_init_gen3()
function, which starts the PHY, yet has no counterpart in the failpath.
Add that counterpart.

Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
Cc: Geert Uytterhoeven <geert+renesas@glider.be>
Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
Cc: Phil Edworthy <phil.edworthy@renesas.com>
Cc: Simon Horman <horms+renesas@verge.net.au>
Cc: Wolfram Sang <wsa@the-dreams.de>
Cc: linux-renesas-soc@vger.kernel.org
Fixes: 517ca93a7159 ("PCI: rcar: Add R-Car gen3 PHY support")
---
V4: New patch
---
 drivers/pci/host/pcie-rcar.c | 10 ++++++++--
 1 file changed, 8 insertions(+), 2 deletions(-)

diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
index 636c3c5095d2..695781934f0a 100644
--- a/drivers/pci/host/pcie-rcar.c
+++ b/drivers/pci/host/pcie-rcar.c
@@ -1163,7 +1163,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 	if (rcar_pcie_hw_init(pcie)) {
 		dev_info(dev, "PCIe link down\n");
 		err = -ENODEV;
-		goto err_clk_disable;
+		goto err_phy_shutdown;
 	}
 
 	data = rcar_pci_read_reg(pcie, MACSR);
@@ -1175,7 +1175,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 			dev_err(dev,
 				"failed to enable MSI support: %d\n",
 				err);
-			goto err_clk_disable;
+			goto err_phy_shutdown;
 		}
 	}
 
@@ -1189,6 +1189,12 @@ static int rcar_pcie_probe(struct platform_device *pdev)
 	if (IS_ENABLED(CONFIG_PCI_MSI))
 		rcar_pcie_teardown_msi(pcie);
 
+err_phy_shutdown:
+	if (pcie->phy) {
+		phy_power_off(pcie->phy);
+		phy_exit(pcie->phy);
+	}
+
 err_clk_disable:
 	clk_disable_unprepare(pcie->bus_clk);
 
-- 
2.16.2

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Lorenzo Pieralisi]

Geert, Simon,

Can I retain your review tags on this series so that I can queue
the patches ? I already added them apart from the last patch that
is new, please let me know asap.

Thanks,
Lorenzo

On Thu, May 24, 2018 at 04:36:18PM +0200, Marek Vasut wrote:
> Multiple minor failpath fixes for the R-Car PCIe driver.
> 
> V4: Sync up the version numbers
>     Rebase on top of Lorenzo's tree
>     Add new patch fixing bug in the PHY code
> 
> Marek Vasut (6):
>   PCI: rcar: Pull bus clock enable/disable from
>     rcar_pcie_get_resources()
>   PCI: rcar: Add missing irq_dispose_mapping() into failpath
>   PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails
>   PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()
>   PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath
>   PCI: rcar: Shut the PHY down in failpath
> 
>  drivers/pci/host/pcie-rcar.c | 82 +++++++++++++++++++++++++++++++++++---------
>  1 file changed, 66 insertions(+), 16 deletions(-)
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org
> 
> -- 
> 2.16.2
> 

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Geert Uytterhoeven]

Hi Lorenzo,

On Fri, May 25, 2018 at 11:35 AM, Lorenzo Pieralisi
<lorenzo.pieralisi@arm.com> wrote:
> Can I retain your review tags on this series so that I can queue
> the patches ? I already added them apart from the last patch that
> is new, please let me know asap.

Yes, please.

Thanks!

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Lorenzo Pieralisi]

On Fri, May 25, 2018 at 11:39:08AM +0200, Geert Uytterhoeven wrote:
> Hi Lorenzo,
> 
> On Fri, May 25, 2018 at 11:35 AM, Lorenzo Pieralisi
> <lorenzo.pieralisi@arm.com> wrote:
> > Can I retain your review tags on this series so that I can queue
> > the patches ? I already added them apart from the last patch that
> > is new, please let me know asap.
> 
> Yes, please.

I assume that's valid also for the last patch, I need Simon's ACK
to proceed too.

Lorenzo

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Geert Uytterhoeven]

Hi Lorenzo,

On Fri, May 25, 2018 at 11:56 AM, Lorenzo Pieralisi
<lorenzo.pieralisi@arm.com> wrote:
> On Fri, May 25, 2018 at 11:39:08AM +0200, Geert Uytterhoeven wrote:
>> On Fri, May 25, 2018 at 11:35 AM, Lorenzo Pieralisi
>> <lorenzo.pieralisi@arm.com> wrote:
>> > Can I retain your review tags on this series so that I can queue
>> > the patches ? I already added them apart from the last patch that
>> > is new, please let me know asap.
>>
>> Yes, please.
>
> I assume that's valid also for the last patch, I need Simon's ACK
> to proceed too.

I haven't reviewed the last patch yet.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH v4 6/6] PCI: rcar: Shut the PHY down in failpath

[Geert Uytterhoeven]

On Thu, May 24, 2018 at 4:36 PM, Marek Vasut <marek.vasut@gmail.com> wrote:
> If anything fails past phy_init_fn() and the system is a Gen3 with
> a PHY, the PHY will be left on and inited. This is caused by the
> phy_init_fn, which is in fact a pointer to rcar_pcie_phy_init_gen3()
> function, which starts the PHY, yet has no counterpart in the failpath.
> Add that counterpart.
>
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>

Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>

Note that there's another unhandled failure mode: rcar_pcie_phy_init_gen3()
doesn't call phy_exit() if phy_power_on() fails.

Gr{oetje,eeting}s,

                        Geert

-- 
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- geert@linux-m68k.org

In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
                                -- Linus Torvalds

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Lorenzo Pieralisi]

On Thu, May 24, 2018 at 04:36:18PM +0200, Marek Vasut wrote:
> Multiple minor failpath fixes for the R-Car PCIe driver.
> 
> V4: Sync up the version numbers
>     Rebase on top of Lorenzo's tree
>     Add new patch fixing bug in the PHY code
> 
> Marek Vasut (6):
>   PCI: rcar: Pull bus clock enable/disable from
>     rcar_pcie_get_resources()
>   PCI: rcar: Add missing irq_dispose_mapping() into failpath
>   PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails
>   PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()
>   PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath
>   PCI: rcar: Shut the PHY down in failpath
> 
>  drivers/pci/host/pcie-rcar.c | 82 +++++++++++++++++++++++++++++++++++---------
>  1 file changed, 66 insertions(+), 16 deletions(-)

Marek,

I have applied patches [1-5] to pci/rcar for v4.18, I could not
apply patch 6 since it is missing Simon's ACK (please have a look
at what I queued to check it is OK).

Bjorn will be merging the branch into -next shortly, I do not know if
there is room for adding additional patches on top of it, as I mentioned
on linux-pci I will be offline for two weeks so it is possible that
what's in pci/rcar at present will be what gets into v4.18, apologies,
that's all I can do.

Lorenzo

> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org
> 
> -- 
> 2.16.2
> 

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Marek Vasut]

On 05/25/2018 06:43 PM, Lorenzo Pieralisi wrote:
> On Thu, May 24, 2018 at 04:36:18PM +0200, Marek Vasut wrote:
>> Multiple minor failpath fixes for the R-Car PCIe driver.
>>
>> V4: Sync up the version numbers
>>     Rebase on top of Lorenzo's tree
>>     Add new patch fixing bug in the PHY code
>>
>> Marek Vasut (6):
>>   PCI: rcar: Pull bus clock enable/disable from
>>     rcar_pcie_get_resources()
>>   PCI: rcar: Add missing irq_dispose_mapping() into failpath
>>   PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails
>>   PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()
>>   PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath
>>   PCI: rcar: Shut the PHY down in failpath
>>
>>  drivers/pci/host/pcie-rcar.c | 82 +++++++++++++++++++++++++++++++++++---------
>>  1 file changed, 66 insertions(+), 16 deletions(-)
> 
> Marek,
> 
> I have applied patches [1-5] to pci/rcar for v4.18, I could not
> apply patch 6 since it is missing Simon's ACK (please have a look
> at what I queued to check it is OK).
> 
> Bjorn will be merging the branch into -next shortly, I do not know if
> there is room for adding additional patches on top of it, as I mentioned
> on linux-pci I will be offline for two weeks so it is possible that
> what's in pci/rcar at present will be what gets into v4.18, apologies,
> that's all I can do.

Fine. Yes, there's another patch coming for the PHY stuff from Sergej.

-- 
Best regards,
Marek Vasut

Re: [PATCH v4 6/6] PCI: rcar: Shut the PHY down in failpath

[Marek Vasut]

On 05/25/2018 12:09 PM, Geert Uytterhoeven wrote:
> On Thu, May 24, 2018 at 4:36 PM, Marek Vasut <marek.vasut@gmail.com> wrote:
>> If anything fails past phy_init_fn() and the system is a Gen3 with
>> a PHY, the PHY will be left on and inited. This is caused by the
>> phy_init_fn, which is in fact a pointer to rcar_pcie_phy_init_gen3()
>> function, which starts the PHY, yet has no counterpart in the failpath.
>> Add that counterpart.
>>
>> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> 
> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be>
> 
> Note that there's another unhandled failure mode: rcar_pcie_phy_init_gen3()
> doesn't call phy_exit() if phy_power_on() fails.

Should be fixed by subsequent patch, thanks.

-- 
Best regards,
Marek Vasut

Re: [PATCH v4 6/6] PCI: rcar: Shut the PHY down in failpath

[Simon Horman]

On Thu, May 24, 2018 at 04:36:24PM +0200, Marek Vasut wrote:
> If anything fails past phy_init_fn() and the system is a Gen3 with
> a PHY, the PHY will be left on and inited. This is caused by the
> phy_init_fn, which is in fact a pointer to rcar_pcie_phy_init_gen3()
> function, which starts the PHY, yet has no counterpart in the failpath.
> Add that counterpart.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org
> Fixes: 517ca93a7159 ("PCI: rcar: Add R-Car gen3 PHY support")

Acked-by: Simon Horman <horms+renesas@verge.net.au>

> ---
> V4: New patch
> ---
>  drivers/pci/host/pcie-rcar.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)
> 
> diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
> index 636c3c5095d2..695781934f0a 100644
> --- a/drivers/pci/host/pcie-rcar.c
> +++ b/drivers/pci/host/pcie-rcar.c
> @@ -1163,7 +1163,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  	if (rcar_pcie_hw_init(pcie)) {
>  		dev_info(dev, "PCIe link down\n");
>  		err = -ENODEV;
> -		goto err_clk_disable;
> +		goto err_phy_shutdown;
>  	}
>  
>  	data = rcar_pci_read_reg(pcie, MACSR);
> @@ -1175,7 +1175,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  			dev_err(dev,
>  				"failed to enable MSI support: %d\n",
>  				err);
> -			goto err_clk_disable;
> +			goto err_phy_shutdown;
>  		}
>  	}
>  
> @@ -1189,6 +1189,12 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  	if (IS_ENABLED(CONFIG_PCI_MSI))
>  		rcar_pcie_teardown_msi(pcie);
>  
> +err_phy_shutdown:
> +	if (pcie->phy) {
> +		phy_power_off(pcie->phy);
> +		phy_exit(pcie->phy);
> +	}
> +
>  err_clk_disable:
>  	clk_disable_unprepare(pcie->bus_clk);
>  
> -- 
> 2.16.2
> 

Re: [PATCH v4 1/6] PCI: rcar: Pull bus clock enable/disable from rcar_pcie_get_resources()

[Simon Horman]

On Thu, May 24, 2018 at 04:36:19PM +0200, Marek Vasut wrote:
> The rcar_pcie_get_resources() is another misnomer with a side effect.
> The function does not only get resources, but also enables/disables bus
> clock. This is forgotten in the probe() function though and if anything
> in probe() fails after rcar_pcie_get_resources() is called, the bus
> clock are never disabled.
> 
> This patch pulls the clock handling out of the rcar_pcie_get_resources()
> and enables clock after all the resources were requested. Moreover, this
> patch also always disables the clock in case of failure.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org
> ---
> V2: No change
> V3: No change
> V4: Rebase on top of Lorenzo's pci/rcar, fix up the rebase breakage
>     induced by patches therein


Acked-by: Simon Horman <horms+renesas@verge.net.au>

> ---
>  drivers/pci/host/pcie-rcar.c | 33 ++++++++++++++++-----------------
>  1 file changed, 16 insertions(+), 17 deletions(-)
> 
> diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
> index 8d161563bce8..0879880f7975 100644
> --- a/drivers/pci/host/pcie-rcar.c
> +++ b/drivers/pci/host/pcie-rcar.c
> @@ -946,32 +946,22 @@ static int rcar_pcie_get_resources(struct rcar_pcie *pcie)
>  		dev_err(dev, "cannot get pcie bus clock\n");
>  		return PTR_ERR(pcie->bus_clk);
>  	}
> -	err = clk_prepare_enable(pcie->bus_clk);
> -	if (err)
> -		return err;
>  
>  	i = irq_of_parse_and_map(dev->of_node, 0);
>  	if (!i) {
>  		dev_err(dev, "cannot get platform resources for msi interrupt\n");
> -		err = -ENOENT;
> -		goto err_map_reg;
> +		return -ENOENT;
>  	}
>  	pcie->msi.irq1 = i;
>  
>  	i = irq_of_parse_and_map(dev->of_node, 1);
>  	if (!i) {
>  		dev_err(dev, "cannot get platform resources for msi interrupt\n");
> -		err = -ENOENT;
> -		goto err_map_reg;
> +		return -ENOENT;
>  	}
>  	pcie->msi.irq2 = i;
>  
>  	return 0;
> -
> -err_map_reg:
> -	clk_disable_unprepare(pcie->bus_clk);
> -
> -	return err;
>  }
>  
>  static int rcar_pcie_inbound_ranges(struct rcar_pcie *pcie,
> @@ -1115,22 +1105,28 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  		goto err_pm_put;
>  	}
>  
> +	err = clk_prepare_enable(pcie->bus_clk);
> +	if (err) {
> +		dev_err(dev, "failed to enable bus clock: %d\n", err);
> +		goto err_pm_put;
> +	}
> +
>  	err = rcar_pcie_parse_map_dma_ranges(pcie, dev->of_node);
>  	if (err)
> -		goto err_pm_put;
> +		goto err_clk_disable;
>  
>  	phy_init_fn = of_device_get_match_data(dev);
>  	err = phy_init_fn(pcie);
>  	if (err) {
>  		dev_err(dev, "failed to init PCIe PHY\n");
> -		goto err_pm_put;
> +		goto err_clk_disable;
>  	}
>  
>  	/* Failure to get a link might just be that no cards are inserted */
>  	if (rcar_pcie_hw_init(pcie)) {
>  		dev_info(dev, "PCIe link down\n");
>  		err = -ENODEV;
> -		goto err_pm_put;
> +		goto err_clk_disable;
>  	}
>  
>  	data = rcar_pci_read_reg(pcie, MACSR);
> @@ -1142,16 +1138,19 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  			dev_err(dev,
>  				"failed to enable MSI support: %d\n",
>  				err);
> -			goto err_pm_put;
> +			goto err_clk_disable;
>  		}
>  	}
>  
>  	err = rcar_pcie_enable(pcie);
>  	if (err)
> -		goto err_pm_put;
> +		goto err_clk_disable;
>  
>  	return 0;
>  
> +err_clk_disable:
> +	clk_disable_unprepare(pcie->bus_clk);
> +
>  err_pm_put:
>  	pm_runtime_put(dev);
>  
> -- 
> 2.16.2
> 

Re: [PATCH v4 2/6] PCI: rcar: Add missing irq_dispose_mapping() into failpath

[Simon Horman]

On Thu, May 24, 2018 at 04:36:20PM +0200, Marek Vasut wrote:
> The rcar_pcie_get_resources() is another misnomer with a side effect.
> The function does not only get resources, but also maps MSI IRQs via
> irq_of_parse_and_map(). In case anything fails afterward, the IRQ
> mapping must be disposed through irq_dispose_mapping() which is not
> done.
> 
> This patch handles irq_of_parse_and_map() failures in by disposing
> of the mapping in rcar_pcie_get_resources() as well as in probe.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org


Acked-by: Simon Horman <horms+renesas@verge.net.au>

Re: [PATCH v4 3/6] PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails

[Simon Horman]

On Thu, May 24, 2018 at 04:36:21PM +0200, Marek Vasut wrote:
> If the rcar_pcie_enable() fails and MSIs are enabled, the setup done in
> rcar_pcie_enable_msi() is never undone. Add a function to tear down the
> MSI setup by disabling the MSI handling in the PCIe block, deallocating
> the pages requested for the MSIs and zapping the IRQ mapping.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org

Reviewed-by: Simon Horman <horms+renesas@verge.net.au>

Re: [PATCH v4 4/6] PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()

[Simon Horman]

On Thu, May 24, 2018 at 04:36:22PM +0200, Marek Vasut wrote:
> The data link active signal usually takes ~20 uSec to be asserted,
> poll the bit more often to avoid useless delays in this function.
> Use udelay() instead of usleep() for such a small delay as suggested
> by the timer documentation and because this will be used in atomic
> context later on when the suspend/resume patches land.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org

Acked-by: Simon Horman <horms+renesas@verge.net.au>

Re: [PATCH v4 5/6] PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath

[Simon Horman]

On Thu, May 24, 2018 at 04:36:23PM +0200, Marek Vasut wrote:
> The rcar_pcie_enable_msi() creates IRQ mappings using irq_create_mapping()
> before requesting the IRQs using devm_request_irq(). If devm_request_irq()
> fails for some reason, rcar_pcie_enable_msi() does not remove the mapping.
> 
> Pull out the code for disposing IRQ mappings from rcar_pcie_teardown_msi()
> into a separate function and call it from both rcar_pcie_teardown_msi()
> and rcar_pcie_enable_msi() failpath to remove the mappings correctly.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Reported-by: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org

Acked-by: Simon Horman <horms+renesas@verge.net.au>

Re: [PATCH v4 0/6] PCI: rcar: Failpath fixes

[Simon Horman]

On Fri, May 25, 2018 at 05:43:13PM +0100, Lorenzo Pieralisi wrote:
> On Thu, May 24, 2018 at 04:36:18PM +0200, Marek Vasut wrote:
> > Multiple minor failpath fixes for the R-Car PCIe driver.
> > 
> > V4: Sync up the version numbers
> >     Rebase on top of Lorenzo's tree
> >     Add new patch fixing bug in the PHY code
> > 
> > Marek Vasut (6):
> >   PCI: rcar: Pull bus clock enable/disable from
> >     rcar_pcie_get_resources()
> >   PCI: rcar: Add missing irq_dispose_mapping() into failpath
> >   PCI: rcar: Teardown MSI setup if rcar_pcie_enable() fails
> >   PCI: rcar: Poll more often in rcar_pcie_wait_for_dl()
> >   PCI: rcar: Remove IRQ mappings in rcar_pcie_enable_msi failpath
> >   PCI: rcar: Shut the PHY down in failpath
> > 
> >  drivers/pci/host/pcie-rcar.c | 82 +++++++++++++++++++++++++++++++++++---------
> >  1 file changed, 66 insertions(+), 16 deletions(-)
> 
> Marek,
> 
> I have applied patches [1-5] to pci/rcar for v4.18, I could not
> apply patch 6 since it is missing Simon's ACK (please have a look
> at what I queued to check it is OK).

Sorry for the delay. You should have that now.

If not, for the entire series:

Acked-by: Simon Horman <horms+renesas@verge.net.au>


> 
> Bjorn will be merging the branch into -next shortly, I do not know if
> there is room for adding additional patches on top of it, as I mentioned
> on linux-pci I will be offline for two weeks so it is possible that
> what's in pci/rcar at present will be what gets into v4.18, apologies,
> that's all I can do.

Re: [PATCH v4 6/6] PCI: rcar: Shut the PHY down in failpath

[Marek Vasut]

On 05/28/2018 10:48 AM, Simon Horman wrote:
> On Thu, May 24, 2018 at 04:36:24PM +0200, Marek Vasut wrote:
>> If anything fails past phy_init_fn() and the system is a Gen3 with
>> a PHY, the PHY will be left on and inited. This is caused by the
>> phy_init_fn, which is in fact a pointer to rcar_pcie_phy_init_gen3()
>> function, which starts the PHY, yet has no counterpart in the failpath.
>> Add that counterpart.
>>
>> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
>> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
>> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
>> Cc: Phil Edworthy <phil.edworthy@renesas.com>
>> Cc: Simon Horman <horms+renesas@verge.net.au>
>> Cc: Wolfram Sang <wsa@the-dreams.de>
>> Cc: linux-renesas-soc@vger.kernel.org
>> Fixes: 517ca93a7159 ("PCI: rcar: Add R-Car gen3 PHY support")
> 
> Acked-by: Simon Horman <horms+renesas@verge.net.au>

While I see the first 5 in today's next, I don't see this one there, so
it'd be nice to add it too.

-- 
Best regards,
Marek Vasut

Re: [PATCH v4 6/6] PCI: rcar: Shut the PHY down in failpath

[Lorenzo Pieralisi]

On Thu, May 24, 2018 at 04:36:24PM +0200, Marek Vasut wrote:
> If anything fails past phy_init_fn() and the system is a Gen3 with
> a PHY, the PHY will be left on and inited. This is caused by the
> phy_init_fn, which is in fact a pointer to rcar_pcie_phy_init_gen3()
> function, which starts the PHY, yet has no counterpart in the failpath.
> Add that counterpart.
> 
> Signed-off-by: Marek Vasut <marek.vasut+renesas@gmail.com>
> Cc: Geert Uytterhoeven <geert+renesas@glider.be>
> Cc: Lorenzo Pieralisi <lorenzo.pieralisi@arm.com>
> Cc: Phil Edworthy <phil.edworthy@renesas.com>
> Cc: Simon Horman <horms+renesas@verge.net.au>
> Cc: Wolfram Sang <wsa@the-dreams.de>
> Cc: linux-renesas-soc@vger.kernel.org
> Fixes: 517ca93a7159 ("PCI: rcar: Add R-Car gen3 PHY support")
> ---
> V4: New patch
> ---
>  drivers/pci/host/pcie-rcar.c | 10 ++++++++--
>  1 file changed, 8 insertions(+), 2 deletions(-)

Applied to pci/controller-fixes to be tentatively merged at -rc4,
thanks.

Lorenzo

> diff --git a/drivers/pci/host/pcie-rcar.c b/drivers/pci/host/pcie-rcar.c
> index 636c3c5095d2..695781934f0a 100644
> --- a/drivers/pci/host/pcie-rcar.c
> +++ b/drivers/pci/host/pcie-rcar.c
> @@ -1163,7 +1163,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  	if (rcar_pcie_hw_init(pcie)) {
>  		dev_info(dev, "PCIe link down\n");
>  		err = -ENODEV;
> -		goto err_clk_disable;
> +		goto err_phy_shutdown;
>  	}
>  
>  	data = rcar_pci_read_reg(pcie, MACSR);
> @@ -1175,7 +1175,7 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  			dev_err(dev,
>  				"failed to enable MSI support: %d\n",
>  				err);
> -			goto err_clk_disable;
> +			goto err_phy_shutdown;
>  		}
>  	}
>  
> @@ -1189,6 +1189,12 @@ static int rcar_pcie_probe(struct platform_device *pdev)
>  	if (IS_ENABLED(CONFIG_PCI_MSI))
>  		rcar_pcie_teardown_msi(pcie);
>  
> +err_phy_shutdown:
> +	if (pcie->phy) {
> +		phy_power_off(pcie->phy);
> +		phy_exit(pcie->phy);
> +	}
> +
>  err_clk_disable:
>  	clk_disable_unprepare(pcie->bus_clk);
>  
> -- 
> 2.16.2
>