[PATCH v2 0/3] USB: EHCI: ehci-mv: make HSIC work

[PATCH v2 0/3] USB: EHCI: ehci-mv: make HSIC work

[Lubomir Rintel]

please have a look at the three patches chained to this message. They
are independent of each other and can be applied iny any order.

My main objective was to make HSIC work on MMP3. The PHY patch is
loosely related to that (there's no MMP3 HSIC PHY driver, but the NOP
is sufficient). The last one is a cosmetic thing.

Compared to first submission, the patches are now submitted as series
and the wording of the third one's commit message has been changed.

Thank you,
Lubo

[PATCH v2 1/3] USB: EHCI: ehci-mv: add HSIC support

[Lubomir Rintel]

Some special dance is needed to initialize the HSIC port.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
---
 drivers/usb/host/ehci-mv.c | 11 +++++++++++
 1 file changed, 11 insertions(+)

diff --git a/drivers/usb/host/ehci-mv.c b/drivers/usb/host/ehci-mv.c
index 66ec1fdf9fe7d..21eb32baf9380 100644
--- a/drivers/usb/host/ehci-mv.c
+++ b/drivers/usb/host/ehci-mv.c
@@ -11,6 +11,7 @@
 #include <linux/clk.h>
 #include <linux/err.h>
 #include <linux/usb/otg.h>
+#include <linux/usb/of.h>
 #include <linux/platform_data/mv_usb.h>
 #include <linux/io.h>
 
@@ -67,6 +68,8 @@ static int mv_ehci_reset(struct usb_hcd *hcd)
 {
 	struct device *dev = hcd->self.controller;
 	struct ehci_hcd_mv *ehci_mv = hcd_to_ehci_hcd_mv(hcd);
+	struct ehci_hcd *ehci = hcd_to_ehci(hcd);
+	u32 status;
 	int retval;
 
 	if (ehci_mv == NULL) {
@@ -80,6 +83,14 @@ static int mv_ehci_reset(struct usb_hcd *hcd)
 	if (retval)
 		dev_err(dev, "ehci_setup failed %d\n", retval);
 
+	if (of_usb_get_phy_mode(dev->of_node) == USBPHY_INTERFACE_MODE_HSIC) {
+		status = ehci_readl(ehci, &ehci->regs->port_status[0]);
+		status |= PORT_TEST_FORCE;
+		ehci_writel(ehci, status, &ehci->regs->port_status[0]);
+		status &= ~PORT_TEST_FORCE;
+		ehci_writel(ehci, status, &ehci->regs->port_status[0]);
+	}
+
 	return retval;
 }
 
-- 
2.24.1

[PATCH v2 2/3] USB: EHCI: ehci-mv: make the PHY optional

[Lubomir Rintel]

We may be using a NOP transceiver and those are treated specially by the
USB core and return -ENODEV with devm_phy_get().

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>
---
 drivers/usb/host/ehci-mv.c | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/drivers/usb/host/ehci-mv.c b/drivers/usb/host/ehci-mv.c
index 21eb32baf9380..c8717f06cd8cb 100644
--- a/drivers/usb/host/ehci-mv.c
+++ b/drivers/usb/host/ehci-mv.c
@@ -127,7 +127,7 @@ static int mv_ehci_probe(struct platform_device *pdev)
 		ehci_mv->set_vbus = pdata->set_vbus;
 	}
 
-	ehci_mv->phy = devm_phy_get(&pdev->dev, "usb");
+	ehci_mv->phy = devm_phy_optional_get(&pdev->dev, "usb");
 	if (IS_ERR(ehci_mv->phy)) {
 		retval = PTR_ERR(ehci_mv->phy);
 		if (retval != -EPROBE_DEFER)
-- 
2.24.1

[PATCH v2 3/3] USB: EHCI: ehci-mv: drop pxa_ehci_type and some device IDs

[Lubomir Rintel]

This is merely a cleanup. None of these is used anywhere.

Signed-off-by: Lubomir Rintel <lkundrak@v3.sk>

---
Changes since v1:
- Changed the wording a bit

 drivers/usb/host/ehci-mv.c           | 6 ++----
 include/linux/platform_data/mv_usb.h | 8 --------
 2 files changed, 2 insertions(+), 12 deletions(-)

diff --git a/drivers/usb/host/ehci-mv.c b/drivers/usb/host/ehci-mv.c
index c8717f06cd8cb..91602e3492084 100644
--- a/drivers/usb/host/ehci-mv.c
+++ b/drivers/usb/host/ehci-mv.c
@@ -257,10 +257,8 @@ static int mv_ehci_remove(struct platform_device *pdev)
 MODULE_ALIAS("mv-ehci");
 
 static const struct platform_device_id ehci_id_table[] = {
-	{"pxa-u2oehci", PXA_U2OEHCI},
-	{"pxa-sph", PXA_SPH},
-	{"mmp3-hsic", MMP3_HSIC},
-	{"mmp3-fsic", MMP3_FSIC},
+	{"pxa-u2oehci", 0},
+	{"pxa-sph", 0},
 	{},
 };
 
diff --git a/include/linux/platform_data/mv_usb.h b/include/linux/platform_data/mv_usb.h
index 5376b6d799d5d..20d239c02bf3a 100644
--- a/include/linux/platform_data/mv_usb.h
+++ b/include/linux/platform_data/mv_usb.h
@@ -6,14 +6,6 @@
 #ifndef __MV_PLATFORM_USB_H
 #define __MV_PLATFORM_USB_H
 
-enum pxa_ehci_type {
-	EHCI_UNDEFINED = 0,
-	PXA_U2OEHCI,	/* pxa 168, 9xx */
-	PXA_SPH,	/* pxa 168, 9xx SPH */
-	MMP3_HSIC,	/* mmp3 hsic */
-	MMP3_FSIC,	/* mmp3 fsic */
-};
-
 enum {
 	MV_USB_MODE_OTG,
 	MV_USB_MODE_HOST,
-- 
2.24.1

Re: [PATCH v2 0/3] USB: EHCI: ehci-mv: make HSIC work

[Alan Stern]

On Sat, 21 Dec 2019, Lubomir Rintel wrote:

> please have a look at the three patches chained to this message. They
> are independent of each other and can be applied iny any order.
> 
> My main objective was to make HSIC work on MMP3. The PHY patch is
> loosely related to that (there's no MMP3 HSIC PHY driver, but the NOP
> is sufficient). The last one is a cosmetic thing.
> 
> Compared to first submission, the patches are now submitted as series
> and the wording of the third one's commit message has been changed.

All three patches are okay with me.

Acked-by: Alan Stern <stern@rowland.harvard.edu>

Alan Stern