Allow xhci-plat using a second clock
Subsystems (3)
cover Allow xhci-plat using a second clock
Hello,
The purpose of this series is to allow xhci-plat using a second clock. It is needed on the Armada 7K/8K but could be used by other SoCs.
The first patch is just a fix found while I was working on this feature.
Thanks,
Gregory
Gregory CLEMENT (2):
usb: host: xhci-plat: Remove useless test before clk_disable_unprepare
usb: host: xhci-plat: Fix clock resource by adding a register clock
Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 ++- drivers/usb/host/xhci-plat.c | 39 ++++++++++++++++------ drivers/usb/host/xhci.h | 3 +- 3 files changed, 35 insertions(+), 12 deletions(-)
Signature · 1 line
2.15.1
PATCH 1/2 usb: host: xhci-plat: Remove useless test before clk_disable_unprepare
clk_disable_unprepare() already checks that the clock pointer is valid. No need to test it before calling it.
--- drivers/usb/host/xhci-plat.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
drivers/usb/host/xhci-plat.c +2 -4
@@ -319,8 +319,7 @@ static int xhci_plat_probe(struct platform_device *pdev)usb_put_hcd (xhci->shared_hcd); disable_clk: -if (!IS_ERR (clk)) -clk_disable_unprepare (clk); +clk_disable_unprepare (clk); put_hcd:usb_put_hcd (hcd);
@@ -346,8 +345,7 @@ static int xhci_plat_remove(struct platform_device *dev)usb_remove_hcd (hcd);usb_put_hcd (xhci->shared_hcd); -if (!IS_ERR (clk)) -clk_disable_unprepare (clk); +clk_disable_unprepare (clk);usb_put_hcd (hcd);pm_runtime_set_suspended (&dev->dev);
Signature · 1 line
2.15.1
PATCH 2/2 usb: host: xhci-plat: Fix clock resource by adding a register clock
In reply to Gregory CLEMENTOn Armada 7K/8K we need to explicitly enable the register clock. This clock is optional because not all the SoCs using this IP need it but at least for Armada 7K/8K it is actually mandatory.
The change was done at xhci-plat level and not at a xhci-mvebu.c because, it is expected that other SoC would have this kind of constraint.
The binding documentation is updating accordingly.
--- Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 +++- drivers/usb/host/xhci-plat.c | 33 ++++++++++++++++++---- drivers/usb/host/xhci.h | 3 +- 3 files changed, 33 insertions(+), 8 deletions(-)
Documentation/devicetree/bindings/usb/usb-xhci.txt +4 -1
@@ -27,7 +27,10 @@ Required properties: - interrupts: one XHCI interrupt should be described here. Optional properties: - - clocks: reference to a clock + - clocks: reference to the clocks + - clock-names: mandatory if there is a second clock, in this case + the name must be "core" for the first clock and "reg" for the + second one - usb2-lpm-disable: indicate if we don't want to enable USB2 HW LPM - usb3-lpm-capable: determines if platform is USB3 LPM capable - quirk-broken-port-ped: set if the controller has broken port disable mechanism
drivers/usb/host/xhci-plat.c +27 -6
@@ -157,6 +157,7 @@ static int xhci_plat_probe(struct platform_device *pdev)struct resource *res;struct usb_hcd *hcd;struct clk *clk; +struct clk *reg_clk;int ret;int irq;
@@ -226,17 +227,27 @@ static int xhci_plat_probe(struct platform_device *pdev) hcd->rsrc_len =resource_size (res);/* -* Not all platforms have a clk so it is not an error if the -* clock does not exists. +* Not all platforms have clks so it is not an error if the +* clock do not exist. */ + reg_clk =devm_clk_get (&pdev->dev,"reg" ); +if (!IS_ERR (reg_clk)) { + ret =clk_prepare_enable (reg_clk); +if (ret) + goto put_hcd; + }else if (PTR_ERR (reg_clk) == -EPROBE_DEFER ) { + ret = -EPROBE_DEFER ; + goto put_hcd; + } + clk =devm_clk_get (&pdev->dev,NULL );if (!IS_ERR (clk)) { ret =clk_prepare_enable (clk);if (ret) - goto put_hcd; + goto disable_reg_clk; }else if (PTR_ERR (clk) == -EPROBE_DEFER ) { ret = -EPROBE_DEFER ; - goto put_hcd; + goto disable_reg_clk; } xhci =hcd_to_xhci (hcd);
@@ -252,6 +263,7 @@ static int xhci_plat_probe(struct platform_device *pdev)device_wakeup_enable (hcd->self.controller); xhci->clk = clk; + xhci->reg_clk = reg_clk; xhci->main_hcd = hcd; xhci->shared_hcd =__usb_create_hcd (driver, sysdev, &pdev->dev,dev_name (&pdev->dev), hcd);
@@ -321,6 +333,9 @@ static int xhci_plat_probe(struct platform_device *pdev) disable_clk:clk_disable_unprepare (clk); +disable_reg_clk: +clk_disable_unprepare (reg_clk); + put_hcd:usb_put_hcd (hcd);
@@ -336,6 +351,7 @@ static int xhci_plat_remove(struct platform_device *dev)struct usb_hcd *hcd =platform_get_drvdata (dev);struct xhci_hcd *xhci =hcd_to_xhci (hcd);struct clk *clk = xhci->clk; +struct clk *reg_clk = xhci->reg_clk; xhci->xhc_state |=XHCI_STATE_REMOVING ;
@@ -346,6 +362,7 @@ static int xhci_plat_remove(struct platform_device *dev)usb_put_hcd (xhci->shared_hcd);clk_disable_unprepare (clk); +clk_disable_unprepare (reg_clk);usb_put_hcd (hcd);pm_runtime_set_suspended (&dev->dev);
@@ -370,8 +387,10 @@ static int __maybe_unused xhci_plat_suspend(struct device *dev)*/ ret =xhci_suspend (xhci,device_may_wakeup (dev)); -if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) +if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) {clk_disable_unprepare (xhci->clk); +clk_disable_unprepare (xhci->reg_clk); + }return ret; }
@@ -382,8 +401,10 @@ static int __maybe_unused xhci_plat_resume(struct device *dev)struct xhci_hcd *xhci =hcd_to_xhci (hcd);int ret; -if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) +if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) { +clk_prepare_enable (xhci->reg_clk);clk_prepare_enable (xhci->clk); + } ret =xhci_priv_resume_quirk (hcd);if (ret)
drivers/usb/host/xhci.h +2 -1
@@ -1726,8 +1726,9 @@ struct xhci_hcd {int page_shift;/* msi-x vectors */ int msix_count; -/* optional clock */ +/* optional clocks */ struct clk *clk; +struct clk *reg_clk;/* data structures */ struct xhci_device_context_array *dcbaa;struct xhci_ring *cmd_ring;
Signature · 1 line
2.15.1
Mathias Nyman
maintainerPATCH 2/2 usb: host: xhci-plat: Fix clock resource by adding a register clock
Quoted 22 lines · Gregory CLEMENT
On Armada 7K/8K we need to explicitly enable the register clock. This clock is optional because not all the SoCs using this IP need it but at least for Armada 7K/8K it is actually mandatory.
The change was done at xhci-plat level and not at a
xhci-mvebu.cbecause, it is expected that other SoC would have this kind of constraint.The binding documentation is updating accordingly.
Signed-off-by Gregory CLEMENT <[email protected]>--- Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 +++- drivers/usb/host/xhci-plat.c | 33 ++++++++++++++++++---- drivers/usb/host/xhci.h | 3 +- 3 files changed, 33 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/usb/usb-xhci.txt +0
-0@@ -27,7 +27,10 @@ Required properties: - interrupts: one XHCI interrupt should be described here.
Documentation/devicetree/bindings/usb/usb-xhci.txt Required properties: · line 36
Quoted 10 lines
Documentation/devicetree/bindings/usb/usb-xhci.txt +4
-1Optional properties: - - clocks: reference to a clock + - clocks: reference to the clocks + - clock-names: mandatory if there is a second clock, in this case + the name must be "core" for the first clock and "reg" for the + second one - usb2-lpm-disable: indicate if we don't want to enable USB2 HW LPM - usb3-lpm-capable: determines if platform is USB3 LPM capable - quirk-broken-port-ped: set if the controller has broken port disable mechanism
Would be good to get a Ack or review by Rob Herring for the above
Quoted 45 lines
drivers/usb/host/xhci-plat.c +15
-4@@ -157,6 +157,7 @@ static int xhci_plat_probe(struct platform_device *pdev)struct resource *res;struct usb_hcd *hcd;struct clk *clk; +struct clk *reg_clk;int ret;int irq;@@ -226,17 +227,27 @@ static int xhci_plat_probe(struct platform_device *pdev) hcd->rsrc_len =resource_size (res);/* -* Not all platforms have a clk so it is not an error if the -* clock does not exists. +* Not all platforms have clks so it is not an error if the +* clock do not exist. */ + reg_clk =devm_clk_get (&pdev->dev,"reg" ); +if (!IS_ERR (reg_clk)) { + ret =clk_prepare_enable (reg_clk); +if (ret) + goto put_hcd; + }else if (PTR_ERR (reg_clk) == -EPROBE_DEFER ) { + ret = -EPROBE_DEFER ; + goto put_hcd; + } + clk =devm_clk_get (&pdev->dev,NULL );if (!IS_ERR (clk)) { ret =clk_prepare_enable (clk);if (ret) - goto put_hcd; + goto disable_reg_clk; }else if (PTR_ERR (clk) == -EPROBE_DEFER ) { ret = -EPROBE_DEFER ; - goto put_hcd; + goto disable_reg_clk; } xhci =hcd_to_xhci (hcd);
drivers/usb/host/xhci-plat.c static int xhci_plat_probe(struct platform_device *pdev) · line 268
Quoted 6 lines
drivers/usb/host/xhci-plat.c +1
-0@@ -252,6 +263,7 @@ static int xhci_plat_probe(struct platform_device *pdev)device_wakeup_enable (hcd->self.controller); xhci->clk = clk; + xhci->reg_clk = reg_clk; xhci->main_hcd = hcd; xhci->shared_hcd = __usb_create_
hcd(driver, sysdev, &pdev->dev,
Quoted 57 lines
drivers/usb/host/xhci-plat.c +11
-2dev_name(&pdev->dev), hcd);@@ -321,6 +333,9 @@ static int xhci_plat_probe(struct platform_device *pdev) disable_clk:clk_disable_unprepare (clk); +disable_reg_clk: +clk_disable_unprepare (reg_clk); + put_hcd:usb_put_hcd (hcd);@@ -336,6 +351,7 @@ static int xhci_plat_remove(struct platform_device *dev)struct usb_hcd *hcd =platform_get_drvdata (dev);struct xhci_hcd *xhci =hcd_to_xhci (hcd);struct clk *clk = xhci->clk; +struct clk *reg_clk = xhci->reg_clk; xhci->xhc_state |=XHCI_STATE_REMOVING ;@@ -346,6 +362,7 @@ static int xhci_plat_remove(struct platform_device *dev)usb_put_hcd (xhci->shared_hcd);clk_disable_unprepare (clk); +clk_disable_unprepare (reg_clk);usb_put_hcd (hcd);pm_runtime_set_suspended (&dev->dev);@@ -370,8 +387,10 @@ static int __maybe_unused xhci_plat_suspend(struct device *dev)*/ ret =xhci_suspend (xhci,device_may_wakeup (dev)); -if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) +if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) {clk_disable_unprepare (xhci->clk); +clk_disable_unprepare (xhci->reg_clk); + }return ret; }@@ -382,8 +401,10 @@ static int __maybe_unused xhci_plat_resume(struct device *dev)struct xhci_hcd *xhci =hcd_to_xhci (hcd);int ret; -if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) +if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) { +clk_prepare_enable (xhci->reg_clk);clk_prepare_enable (xhci->clk); + } ret =xhci_priv_resume_quirk (hcd);if (ret)drivers/usb/host/xhci.h +0
-0
drivers/usb/host/xhci.h struct xhci_hcd { · line 1734
Quoted 10 lines
drivers/usb/host/xhci.h +2
-1@@ -1726,8 +1726,9 @@ struct xhci_hcd {int page_shift;/* msi-x vectors */ int msix_count; -/* optional clock */ +/* optional clocks */ struct clk *clk; +struct clk *reg_clk;/* data structures */ struct xhci_device_context_array *dcbaa;struct xhci_ring *cmd_ring;
Looks good to me, I'll queue it up if the devicetree bindings looks ok for Rob Herring
-Mathias
PATCH 2/2 usb: host: xhci-plat: Fix clock resource by adding a register clock
Hi Rob,
Quoted 26 lines · Mathias Nyman
Quoted 25 lines · Gregory CLEMENT
On Armada 7K/8K we need to explicitly enable the register clock. This clock is optional because not all the SoCs using this IP need it but at least for Armada 7K/8K it is actually mandatory.
The change was done at xhci-plat level and not at a
xhci-mvebu.cbecause, it is expected that other SoC would have this kind of constraint.The binding documentation is updating accordingly.
Signed-off-by Gregory CLEMENT <[email protected]>--- Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 +++- drivers/usb/host/xhci-plat.c | 33 ++++++++++++++++++---- drivers/usb/host/xhci.h | 3 +- 3 files changed, 33 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/usb/usb-xhci.txt +1
-1@@ -27,7 +27,10 @@ Required properties: - interrupts: one XHCI interrupt should be described here. Optional properties: - - clocks: reference to a clock + - clocks: reference to the clocks
Quoted 8 lines
Documentation/devicetree/bindings/usb/usb-xhci.txt Required properties: · line 36
Quoted 6 lines
Documentation/devicetree/bindings/usb/usb-xhci.txt +3
-0+ - clock-names: mandatory if there is a second clock, in this case + the name must be "core" for the first clock and "reg" for the + second one - usb2-lpm-disable: indicate if we don't want to enable USB2 HW LPM - usb3-lpm-capable: determines if platform is USB3 LPM capable - quirk-broken-port-ped: set if the controller has broken port disable mechanismWould be good to get a Ack or review by Rob Herring for the above
Would you mind to have a look on this binding update ?
Initially I didn't copy you trying to not overflow you because this change is only about adding a new clock, which is pretty common. But I think Mathias would be more confident with a Ack or review by you.
Thanks,
Gregory
Quoted 115 lines
Quoted 44 lines
drivers/usb/host/xhci-plat.c +1
-0@@ -157,6 +157,7 @@ static int xhci_plat_probe(struct platform_device *pdev)struct resource *res;struct usb_hcd *hcd;struct clk *clk; +struct clk *reg_clk;int ret;int irq; @@ -226,17 +227,27 @@ static int xhci_plat_probe(struct
platform_device*pdev)hcd->rsrc_len = resource_size(res); /*drivers/usb/host/xhci-plat.c +14
-4-* Not all platforms have a clk so it is not an error if the -* clock does not exists. +* Not all platforms have clks so it is not an error if the +* clock do not exist. */ + reg_clk =devm_clk_get (&pdev->dev,"reg" ); +if (!IS_ERR (reg_clk)) { + ret =clk_prepare_enable (reg_clk); +if (ret) + goto put_hcd; + }else if (PTR_ERR (reg_clk) == -EPROBE_DEFER ) { + ret = -EPROBE_DEFER ; + goto put_hcd; + } + clk =devm_clk_get (&pdev->dev,NULL );if (!IS_ERR (clk)) { ret =clk_prepare_enable (clk);if (ret) - goto put_hcd; + goto disable_reg_clk; }else if (PTR_ERR (clk) == -EPROBE_DEFER ) { ret = -EPROBE_DEFER ; - goto put_hcd; + goto disable_reg_clk; } xhci =hcd_to_xhci (hcd);drivers/usb/host/xhci-plat.c static int xhci_plat_probe(struct platform_device *pdev) · line 267
Quoted 5 lines
drivers/usb/host/xhci-plat.c +1
-0@@ -252,6 +263,7 @@ static int xhci_plat_probe(struct platform_device *pdev)device_wakeup_enable (hcd->self.controller); xhci->clk = clk; + xhci->reg_clk = reg_clk; xhci->main_hcd = hcd; xhci->shared_hcd = __usb_create_hcd(driver, sysdev,
&pdev->dev,Quoted 49 lines
drivers/usb/host/xhci-plat.c +2
-0dev_name(&pdev->dev), hcd);@@ -321,6 +333,9 @@ static int xhci_plat_probe(struct platform_device *pdev) disable_clk:clk_disable_unprepare (clk); +disable_reg_clk: +clk_disable_unprepare (reg_clk); + put_hcd:usb_put_hcd (hcd); @@ -336,6 +351,7 @@ static int xhci_plat_remove(struct
platform_device*dev)
structusb_hcd*hcd =platform_get_drvdata(dev);
structxhci_hcd*xhci =hcd_to_xhci(hcd);
struct clk *clk =xhci->clk;drivers/usb/host/xhci-plat.c +1
-0+struct clk *reg_clk = xhci->reg_clk; xhci->xhc_state |=XHCI_STATE_REMOVING ; @@ -346,6 +362,7 @@ static int xhci_plat_remove(struct
platform_device*dev)usb_put_hcd(xhci->shared_hcd); clk_disable_unprepare(clk);drivers/usb/host/xhci-plat.c +7
-0+clk_disable_unprepare (reg_clk);usb_put_hcd (hcd);pm_runtime_set_suspended (&dev->dev);@@ -370,8 +387,10 @@ static int __maybe_unused xhci_plat_suspend(struct device *dev)*/ ret =xhci_suspend (xhci,device_may_wakeup (dev)); - if (!device_may_wakeup(dev) && !IS_ERR(xhci->clk)) +if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) {clk_disable_unprepare (xhci->clk); +clk_disable_unprepare (xhci->reg_clk); + }return ret; }@@ -382,8 +401,10 @@ static int __maybe_unused xhci_plat_resume(struct device *dev)struct xhci_hcd *xhci =hcd_to_xhci (hcd);int ret; - if (!device_may_wakeup(dev) && !IS_ERR(xhci->clk)) +if (!device_may_wakeup (dev) && !IS_ERR (xhci->clk)) { +clk_prepare_enable (xhci->reg_clk);clk_prepare_enable (xhci->clk); + } ret =xhci_priv_resume_quirk (hcd);if (ret)drivers/usb/host/xhci.h +0
-0drivers/usb/host/xhci.h struct xhci_hcd { · line 1734
Quoted 10 lines
drivers/usb/host/xhci.h +2
-1@@ -1726,8 +1726,9 @@ struct xhci_hcd {int page_shift;/* msi-x vectors */ int msix_count; -/* optional clock */ +/* optional clocks */ struct clk *clk; +struct clk *reg_clk;/* data structures */ struct xhci_device_context_array *dcbaa;struct xhci_ring *cmd_ring;Looks good to me, I'll queue it up if the devicetree bindings looks ok for Rob Herring
-Mathias
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering http://bootlin.com
PATCH 2/2 usb: host: xhci-plat: Fix clock resource by adding a register clock
In reply to Mathias NymanHi Rob,
(resent because of malformed cc list)
Quoted 26 lines · Mathias Nyman
Quoted 25 lines · Gregory CLEMENT
On Armada 7K/8K we need to explicitly enable the register clock. This clock is optional because not all the SoCs using this IP need it but at least for Armada 7K/8K it is actually mandatory.
The change was done at xhci-plat level and not at a
xhci-mvebu.cbecause, it is expected that other SoC would have this kind of constraint.The binding documentation is updating accordingly.
Signed-off-by Gregory CLEMENT <[email protected]>--- Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 +++- drivers/usb/host/xhci-plat.c | 33 ++++++++++++++++++---- drivers/usb/host/xhci.h | 3 +- 3 files changed, 33 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/usb/usb-xhci.txt +1
-1@@ -27,7 +27,10 @@ Required properties: - interrupts: one XHCI interrupt should be described here. Optional properties: - - clocks: reference to a clock + - clocks: reference to the clocks
Quoted 8 lines
Documentation/devicetree/bindings/usb/usb-xhci.txt Required properties: · line 36
Quoted 6 lines
Documentation/devicetree/bindings/usb/usb-xhci.txt +3
-0+ - clock-names: mandatory if there is a second clock, in this case + the name must be "core" for the first clock and "reg" for the + second one - usb2-lpm-disable: indicate if we don't want to enable USB2 HW LPM - usb3-lpm-capable: determines if platform is USB3 LPM capable - quirk-broken-port-ped: set if the controller has broken port disable mechanismWould be good to get a Ack or review by Rob Herring for the above
Would you mind to have a look on this binding update ?
Initially I didn't copy you trying to not overflow you because this change is only about adding a new clock, which is pretty common. But I think Mathias would be more confident with a Ack or review by you.
Thanks,
Gregory
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering http://bootlin.com
PATCH 2/2 usb: host: xhci-plat: Fix clock resource by adding a register clock
In reply to Gregory CLEMENTHi,
Quoted 53 lines · Gregory CLEMENT
On Armada 7K/8K we need to explicitly enable the register clock. This clock is optional because not all the SoCs using this IP need it but at least for Armada 7K/8K it is actually mandatory.
The change was done at xhci-plat level and not at a
xhci-mvebu.cbecause, it is expected that other SoC would have this kind of constraint.The binding documentation is updating accordingly.
Signed-off-by Gregory CLEMENT <[email protected]>--- Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 +++- drivers/usb/host/xhci-plat.c | 33 ++++++++++++++++++---- drivers/usb/host/xhci.h | 3 +- 3 files changed, 33 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/usb/usb-xhci.txt +4
-1@@ -27,7 +27,10 @@ Required properties: - interrupts: one XHCI interrupt should be described here. Optional properties: - - clocks: reference to a clock + - clocks: reference to the clocks + - clock-names: mandatory if there is a second clock, in this case + the name must be "core" for the first clock and "reg" for the + second one - usb2-lpm-disable: indicate if we don't want to enable USB2 HW LPM - usb3-lpm-capable: determines if platform is USB3 LPM capable - quirk-broken-port-ped: set if the controller has broken port disable mechanismdrivers/usb/host/xhci-plat.c +3
-2@@ -157,6 +157,7 @@ static int xhci_plat_probe(struct platform_device *pdev)struct resource *res;struct usb_hcd *hcd;struct clk *clk; +struct clk *reg_clk;int ret;int irq;@@ -226,17 +227,27 @@ static int xhci_plat_probe(struct platform_device *pdev) hcd->rsrc_len =resource_size (res);/* -* Not all platforms have a clk so it is not an error if the -* clock does not exists. +* Not all platforms have clks so it is not an error if the +* clock do not exist. */
drivers/usb/host/xhci-plat.c static int xhci_plat_probe(struct platform_device *pdev) · line 242
Quoted 10 lines
drivers/usb/host/xhci-plat.c +10
-0+ reg_clk =devm_clk_get (&pdev->dev,"reg" ); +if (!IS_ERR (reg_clk)) { + ret =clk_prepare_enable (reg_clk); +if (ret) + goto put_hcd; + }else if (PTR_ERR (reg_clk) == -EPROBE_DEFER ) { + ret = -EPROBE_DEFER ; + goto put_hcd; + } +
How about using clk_bulk_ APIs?
Quoted 4 lines
drivers/usb/host/xhci-plat.c +0
-0clk =devm_clk_get (&pdev->dev,NULL );if (!IS_ERR (clk)) { ret =clk_prepare_enable (clk);
Signature · 2 lines
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project
PATCH 2/2 usb: host: xhci-plat: Fix clock resource by adding a register clock
Hi Manu,
Quoted 61 lines · Manu Gautam
Hi,
Quoted 57 lines · Gregory CLEMENT
On Armada 7K/8K we need to explicitly enable the register clock. This clock is optional because not all the SoCs using this IP need it but at least for Armada 7K/8K it is actually mandatory.
The change was done at xhci-plat level and not at a
xhci-mvebu.cbecause, it is expected that other SoC would have this kind of constraint.The binding documentation is updating accordingly.
Signed-off-by Gregory CLEMENT <[email protected]>--- Documentation/devicetree/bindings/usb/usb-xhci.txt | 5 +++- drivers/usb/host/xhci-plat.c | 33 ++++++++++++++++++---- drivers/usb/host/xhci.h | 3 +- 3 files changed, 33 insertions(+), 8 deletions(-)Documentation/devicetree/bindings/usb/usb-xhci.txt +4
-1@@ -27,7 +27,10 @@ Required properties: - interrupts: one XHCI interrupt should be described here. Optional properties: - - clocks: reference to a clock + - clocks: reference to the clocks + - clock-names: mandatory if there is a second clock, in this case + the name must be "core" for the first clock and "reg" for the + second one - usb2-lpm-disable: indicate if we don't want to enable USB2 HW LPM - usb3-lpm-capable: determines if platform is USB3 LPM capable - quirk-broken-port-ped: set if the controller has broken port disable mechanismdrivers/usb/host/xhci-plat.c +7
-2@@ -157,6 +157,7 @@ static int xhci_plat_probe(struct platform_device *pdev)struct resource *res;struct usb_hcd *hcd;struct clk *clk; +struct clk *reg_clk;int ret;int irq;@@ -226,17 +227,27 @@ static int xhci_plat_probe(struct platform_device *pdev) hcd->rsrc_len =resource_size (res);/* -* Not all platforms have a clk so it is not an error if the -* clock does not exists. +* Not all platforms have clks so it is not an error if the +* clock do not exist. */ + reg_clk =devm_clk_get (&pdev->dev,"reg" ); +if (!IS_ERR (reg_clk)) { + ret =clk_prepare_enable (reg_clk); +if (ret)
Quoted 8 lines
drivers/usb/host/xhci-plat.c static int xhci_plat_probe(struct platform_device *pdev) · line 242
Quoted 6 lines
drivers/usb/host/xhci-plat.c +6
-0+ goto put_hcd; + }else if (PTR_ERR (reg_clk) == -EPROBE_DEFER ) { + ret = -EPROBE_DEFER ; + goto put_hcd; + } +How about using
clk_bulk_APIs?
I didn't know this API, but after having a look on it, it didn't match what I need. Indeed the second clock is "optional" to handle the backward compatibility. With the clk_bulk_ APIS all the clocks are mandatory.
Thanks,
Gregory
Quoted 15 lines
Quoted 5 lines
drivers/usb/host/xhci-plat.c +0
-0clk =devm_clk_get (&pdev->dev,NULL );if (!IS_ERR (clk)) { ret =clk_prepare_enable (clk);Signature · 8 lines
The Qualcomm Innovation Center, Inc. is a member of the Code Aurora Forum, a Linux Foundation Collaborative Project _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel at lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering http://bootlin.com
cover Allow xhci-plat using a second clock
In reply to Gregory CLEMENTHi Mathias,
Quoted 5 lines · Gregory CLEMENT
Hello,
The purpose of this series is to allow xhci-plat using a second clock. It is needed on the Armada 7K/8K but could be used by other SoCs.
Do you have some comments on this series and especially on the second patch?
Thanks,
Gregory
Quoted 21 lines
The first patch is just a fix found while I was working on this feature.
Thanks,
Gregory
Gregory CLEMENT (2):
usb: host: xhci-plat: Remove useless test beforeclk_disable_unprepare
usb: host: xhci-plat: Fix clock resource by adding a register clockDocumentation/devicetree/bindings/usb/usb-xhci.txt | 5 ++- drivers/usb/host/xhci-plat.c | 39 ++++++++++++++++------ drivers/usb/host/xhci.h | 3 +- 3 files changed, 35 insertions(+), 12 deletions(-)Signature · 1 line
2.15.1
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering http://bootlin.com
cover Allow xhci-plat using a second clock
In reply to Gregory CLEMENTHi Mathias,
Quoted 8 lines · Gregory CLEMENT
Hello,
The purpose of this series is to allow xhci-plat using a second clock. It is needed on the Armada 7K/8K but could be used by other SoCs.
The first patch is just a fix found while I was working on this feature.
I've just realized that this series sent 2 months ago was not merged in v4.17. The issue is that now the USB support on the Armada 7K/8K is broken, because the clock support part was already merged.
You already had a look on this series one month ago. The issue you had was about getting an approval for the extension of the binding [1]. I pinged Rob about it [2] one month ago but we didn't get any feedback. However, Rob already approved the similar changes I introduced in an other for mv_xor_v2: [3].
So would it be possible to apply this series on v4.17-rc ?
Thanks,
Gregory
[1]:
http://lists.infradead.org/pipermail/linux-arm-kernel/2018-February/562739.html
[2]:
http://lists.infradead.org/pipermail/linux-arm-kernel/2018-March/566076.html
[3]:
http://lists.infradead.org/pipermail/linux-arm-kernel/2018-March/564594.html
Quoted 23 lines
Thanks,
Gregory
Gregory CLEMENT (2):
usb: host: xhci-plat: Remove useless test beforeclk_disable_unprepare
usb: host: xhci-plat: Fix clock resource by adding a register clockDocumentation/devicetree/bindings/usb/usb-xhci.txt | 5 ++- drivers/usb/host/xhci-plat.c | 39 ++++++++++++++++------ drivers/usb/host/xhci.h | 3 +- 3 files changed, 35 insertions(+), 12 deletions(-)Signature · 7 lines
2.15.1 _______________________________________________ linux-arm-kernel mailing list linux-arm-kernel at lists.infradead.org http://lists.infradead.org/mailman/listinfo/linux-arm-kernel
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering http://bootlin.com
cover Allow xhci-plat using a second clock
Quoted 17 lines · Gregory CLEMENT
Hi Mathias,
Quoted 8 lines · Gregory CLEMENT
Hello,
The purpose of this series is to allow xhci-plat using a second clock. It is needed on the Armada 7K/8K but could be used by other SoCs.
The first patch is just a fix found while I was working on this feature.
I've just realized that this series sent 2 months ago was not merged in v4.17. The issue is that now the USB support on the Armada 7K/8K is broken, because the clock support part was already merged.
Quoted 7 lines
You already had a look on this series one month ago. The issue you had was about getting an approval for the extension of the binding [1]. I pinged Rob about it [2] one month ago but we didn't get any feedback. However, Rob already approved the similar changes I introduced in an other for
mv_xor_v2: [3].So would it be possible to apply this series on v4.17-rc ?
Patch 2/2 no longer applies due to clk suspend/resume changes in:
commit d56e57ca030c8b4296944a2ae61ac167bf979c07
usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"
Not sure how that change affects Armada 7K/8K
Can you rebase and resend the series?
-Mathias
cover Allow xhci-plat using a second clock
Hi Mathias,
Quoted 26 lines · Mathias Nyman
Quoted 17 lines · Gregory CLEMENT
Hi Mathias,
On mer., f?vr. 14 2018, Gregory CLEMENTQuoted 8 lines · <[email protected]> wrote:
Hello,
The purpose of this series is to allow xhci-plat using a second clock. It is needed on the Armada 7K/8K but could be used by other SoCs.
The first patch is just a fix found while I was working on this feature.
I've just realized that this series sent 2 months ago was not merged in v4.17. The issue is that now the USB support on the Armada 7K/8K is broken, because the clock support part was already merged.
Quoted 7 lines
You already had a look on this series one month ago. The issue you had was about getting an approval for the extension of the binding [1]. I pinged Rob about it [2] one month ago but we didn't get any feedback. However, Rob already approved the similar changes I introduced in an other for
mv_xor_v2: [3].So would it be possible to apply this series on v4.17-rc ?
Quoted 7 lines
Patch 2/2 no longer applies due to clk suspend/resume changes in:
commit
d56e57ca030c8b4296944a2ae61ac167bf979c07
usb: host: xhci-plat: revert "usb: host: xhci-plat: enable clk in resume timing"Not sure how that change affects Armada 7K/8K
Can you rebase and resend the series?
It's done!
About the suspend/resume part it does not affect Armada 7K/8K because we don't support it yet.
However I am a bit surprised by the commit log because reverting a common part because of a specific implementation looks wrong. I expect issues when we will add PM support but we will see that later.
Gregory
Quoted 3 lines
-Mathias
Signature · 3 lines
Gregory Clement, Bootlin (formerly Free Electrons) Embedded Linux and Kernel engineering http://bootlin.com