Skip to contentSkip to conversation

usb: dwc2: Add support for dual-role

PATCH Superseded [email protected] 4 participants 32 messages Oct 28, 2014 – Nov 3, 2014

Subsystems (2)

v6 of 7 Superseded Latest is v7 →

lore
Share & tools
lore
#0 lore raw

v6 cover usb: dwc2: Add support for dual-role

From: Dinh Nguyen <[email protected]>

Hello,

This is version 6 of the patch series that combines the dwc2 gadget and host driver into a single dual role driver. Here are the main differences from V5:

- Added a new patch 8/8 - usb: dwc2: move usb_disabled() call to host driver only. This patch is needed to avoid a build error when (!USB && USB_GADGET) condition is met.
- Addressed comments in the Kconfig/Makefile. USB_DWC2_PLATFORM and USB_DWC2_PCI are now tristate.

For v6, the series is rebased on top of Felipe Balbi's tree on testing/next branch. I thought this might be appropriate as there are dwc2 patches already on this branch.

As usual, tested on SOCFPGA(host, gadget, and dual-role) and on Rpi-B (host mode only).

I have pushed this series to a git repo to make it more convenient for people to test/review.

git://git.rocketboards.org/linux-socfpga-next.git dwc2_dual_role_v6

Thanks,

Dinh Nguyen (8):
usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure
usb: dwc2: Move gadget probe function into platform code
usb: dwc2: Initialize the USB core for peripheral mode
usb: dwc2: Update common interrupt handler to call gadget interrupt
handler
usb: dwc2: Add call_gadget functions for perpheral mode interrupts
usb: dwc2: gadget: Do not fail probe if there isn't a clock node
usb: dwc2: Update Kconfig to support dual-role
usb: dwc2: move usb_disabled() call to host driver only

 drivers/usb/dwc2/Kconfig     |  66 +++++-----
 drivers/usb/dwc2/Makefile    |  32 ++---
 drivers/usb/dwc2/core.c      |  10 --
 drivers/usb/dwc2/core.h      | 192 +++++++++++++++++------------
 drivers/usb/dwc2/core_intr.c |  16 ++-
 drivers/usb/dwc2/gadget.c    | 283 +++++++++++++------------------------------
 drivers/usb/dwc2/hcd.c       |   6 +-
 drivers/usb/dwc2/hcd.h       |  10 --
 drivers/usb/dwc2/pci.c       |   7 ++
 drivers/usb/dwc2/platform.c  |  55 ++++++++-
 10 files changed, 336 insertions(+), 341 deletions(-)
Signature · 1 line
2.0.3
#1 lore raw

PATCH v6 1/8 usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure

From: Dinh Nguyen <[email protected]>

Adds the gadget data structure and appropriate data structure pointers to the common dwc2_hsotg data structure. To keep the driver data dereference code looking clean, the gadget variable declares are only available for peripheral and dual-role mode. This is needed so that the dwc2_hsotg data structure can be used by the hcd and gadget drivers.

Updates gadget.c to use the dwc2_hsotg data structure and gadget pointers that have been moved into the common dwc2_hsotg structure.

---
v5: Keep the changes to mininum and maintain hcd and gadget driver to build
and work separately. Use IS_ENABLED() instead of #if defined
v3: Updated with paulz's suggestion to avoid double pointers.
v2: Left the function parameter name as 'hsotg' and just changed its type.

---
 drivers/usb/dwc2/core.h   | 156 ++++++++++++++++++++++++----------------------
 drivers/usb/dwc2/gadget.c | 145 +++++++++++++++++++++---------------------
 2 files changed, 154 insertions(+), 147 deletions(-)
drivers/usb/dwc2/core.h +81 -75
index 55c90c5..96c283d 100644
@@ -84,7 +84,7 @@ static const char * const s3c_hsotg_supply_names[] = {
  */
 #define EP0_MPS_LIMIT   64
 
-struct s3c_hsotg;
+struct dwc2_hsotg;
 struct s3c_hsotg_req;
 
 /**
@@ -130,7 +130,7 @@ struct s3c_hsotg_req;
 struct s3c_hsotg_ep {
 	struct usb_ep           ep;
 	struct list_head        queue;
-	struct s3c_hsotg        *parent;
+	struct dwc2_hsotg       *parent;
 	struct s3c_hsotg_req    *req;
 	struct dentry           *debugfs;
 
@@ -155,67 +155,6 @@ struct s3c_hsotg_ep {
 };
 
 /**
- * struct s3c_hsotg - driver state.
- * @dev: The parent device supplied to the probe function
- * @driver: USB gadget driver
- * @phy: The otg phy transceiver structure for phy control.
- * @uphy: The otg phy transceiver structure for old USB phy control.
- * @plat: The platform specific configuration data. This can be removed once
- * all SoCs support usb transceiver.
- * @regs: The memory area mapped for accessing registers.
- * @irq: The IRQ number we are using
- * @supplies: Definition of USB power supplies
- * @phyif: PHY interface width
- * @dedicated_fifos: Set if the hardware has dedicated IN-EP fifos.
- * @num_of_eps: Number of available EPs (excluding EP0)
- * @debug_root: root directrory for debugfs.
- * @debug_file: main status file for debugfs.
- * @debug_fifo: FIFO status file for debugfs.
- * @ep0_reply: Request used for ep0 reply.
- * @ep0_buff: Buffer for EP0 reply data, if needed.
- * @ctrl_buff: Buffer for EP0 control requests.
- * @ctrl_req: Request for EP0 control packets.
- * @setup: NAK management for EP0 SETUP
- * @last_rst: Time of last reset
- * @eps: The endpoints being supplied to the gadget framework
- */
-struct s3c_hsotg {
-	struct device            *dev;
-	struct usb_gadget_driver *driver;
-	struct phy               *phy;
-	struct usb_phy           *uphy;
-	struct s3c_hsotg_plat    *plat;
-
-	spinlock_t              lock;
-
-	void __iomem            *regs;
-	int                     irq;
-	struct clk              *clk;
-
-	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
-
-	u32                     phyif;
-	int			fifo_mem;
-	unsigned int            dedicated_fifos:1;
-	unsigned char           num_of_eps;
-	u32			fifo_map;
-
-	struct dentry           *debug_root;
-	struct dentry           *debug_file;
-	struct dentry           *debug_fifo;
-
-	struct usb_request      *ep0_reply;
-	struct usb_request      *ctrl_req;
-	u8                      ep0_buff[8];
-	u8                      ctrl_buff[8];
-
-	struct usb_gadget       gadget;
-	unsigned int            setup;
-	unsigned long           last_rst;
-	struct s3c_hsotg_ep     *eps;
-};
-
-/**
  * struct s3c_hsotg_req - data transfer request
  * @req: The USB gadget request
  * @queue: The list of requests for the endpoint this is queued for.
@@ -229,6 +168,7 @@ struct s3c_hsotg_req {
 	unsigned char           mapped;
 };
 
+#if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 #define call_gadget(_hs, _entry) \
 do { \
 	if ((_hs)->gadget.speed != USB_SPEED_UNKNOWN && \
@@ -238,6 +178,9 @@ do { \
 		spin_lock(&_hs->lock); \
 	} \
 } while (0)
+#else
+#define call_gadget(_hs, _entry)	do {} while (0)
+#endif
 
 struct dwc2_hsotg;
 struct dwc2_host_chan;
@@ -495,11 +438,13 @@ struct dwc2_hw_params {
  * struct dwc2_hsotg - Holds the state of the driver, including the non-periodic
  * and periodic schedules
  *
+ * These are common for both host and peripheral modes:
+ *
  * @dev:                The struct device pointer
  * @regs:		Pointer to controller regs
- * @core_params:        Parameters that define how the core should be configured
  * @hw_params:          Parameters that were autodetected from the
  *                      hardware registers
+ * @core_params:	Parameters that define how the core should be configured
  * @op_state:           The operational State, during transitions (a_host=>
  *                      a_peripheral and b_device=>b_host) this may not match
  *                      the core, but allows the software to determine
@@ -508,6 +453,8 @@ struct dwc2_hw_params {
  *                      - USB_DR_MODE_PERIPHERAL
  *                      - USB_DR_MODE_HOST
  *                      - USB_DR_MODE_OTG
+ * @lock:		Spinlock that protects all the driver data structures
+ * @priv:		Stores a pointer to the struct usb_hcd
  * @queuing_high_bandwidth: True if multiple packets of a high-bandwidth
  *                      transfer are in process of being queued
  * @srp_success:        Stores status of SRP request in the case of a FS PHY
@@ -517,6 +464,9 @@ struct dwc2_hw_params {
  *                      interrupt
  * @wkp_timer:          Timer object for handling Wakeup Detected interrupt
  * @lx_state:           Lx state of connected device
+ *
+ * These are for host mode:
+ *
  * @flags:              Flags for handling root port state changes
  * @non_periodic_sched_inactive: Inactive QHs in the non-periodic schedule.
  *                      Transfers associated with these QHs are not currently
@@ -585,11 +535,31 @@ struct dwc2_hw_params {
  * @status_buf_dma:     DMA address for status_buf
  * @start_work:         Delayed work for handling host A-cable connection
  * @reset_work:         Delayed work for handling a port reset
- * @lock:               Spinlock that protects all the driver data structures
- * @priv:               Stores a pointer to the struct usb_hcd
  * @otg_port:           OTG port number
  * @frame_list:         Frame list
  * @frame_list_dma:     Frame list DMA address
+ *
+ * These are for peripheral mode:
+ *
+ * @driver:             USB gadget driver
+ * @phy:                The otg phy transceiver structure for phy control.
+ * @uphy:               The otg phy transceiver structure for old USB phy control.
+ * @plat:               The platform specific configuration data. This can be removed once
+ *                      all SoCs support usb transceiver.
+ * @supplies:           Definition of USB power supplies
+ * @phyif:              PHY interface width
+ * @dedicated_fifos:    Set if the hardware has dedicated IN-EP fifos.
+ * @num_of_eps:         Number of available EPs (excluding EP0)
+ * @debug_root:         Root directrory for debugfs.
+ * @debug_file:         Main status file for debugfs.
+ * @debug_fifo:         FIFO status file for debugfs.
+ * @ep0_reply:          Request used for ep0 reply.
+ * @ep0_buff:           Buffer for EP0 reply data, if needed.
+ * @ctrl_buff:          Buffer for EP0 control requests.
+ * @ctrl_req:           Request for EP0 control packets.
+ * @setup:              NAK management for EP0 SETUP
+ * @last_rst:           Time of last reset
+ * @eps:                The endpoints being supplied to the gadget framework
  */
 struct dwc2_hsotg {
 	struct device *dev;
@@ -601,6 +571,9 @@ struct dwc2_hsotg {
 	enum usb_otg_state op_state;
 	enum usb_dr_mode dr_mode;
 
+	spinlock_t lock;
+	void *priv;
+
 	unsigned int queuing_high_bandwidth:1;
 	unsigned int srp_success:1;
 
@@ -609,6 +582,14 @@ struct dwc2_hsotg {
 	struct timer_list wkp_timer;
 	enum dwc2_lx_state lx_state;
 
+	/* DWC OTG HW Release versions */
+#define DWC2_CORE_REV_2_71a	0x4f54271a
+#define DWC2_CORE_REV_2_90a	0x4f54290a
+#define DWC2_CORE_REV_2_92a	0x4f54292a
+#define DWC2_CORE_REV_2_94a	0x4f54294a
+#define DWC2_CORE_REV_3_00a	0x4f54300a
+
+#if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 	union dwc2_hcd_internal_flags {
 		u32 d32;
 		struct {
@@ -655,19 +636,10 @@ struct dwc2_hsotg {
 
 	struct delayed_work start_work;
 	struct delayed_work reset_work;
-	spinlock_t lock;
-	void *priv;
 	u8 otg_port;
 	u32 *frame_list;
 	dma_addr_t frame_list_dma;
 
-	/* DWC OTG HW Release versions */
-#define DWC2_CORE_REV_2_71a	0x4f54271a
-#define DWC2_CORE_REV_2_90a	0x4f54290a
-#define DWC2_CORE_REV_2_92a	0x4f54292a
-#define DWC2_CORE_REV_2_94a	0x4f54294a
-#define DWC2_CORE_REV_3_00a	0x4f54300a
-
 #ifdef DEBUG
 	u32 frrem_samples;
 	u64 frrem_accum;
@@ -686,6 +658,40 @@ struct dwc2_hsotg {
 	u32 hfnum_other_samples_b;
 	u64 hfnum_other_frrem_accum_b;
 #endif
+#endif /* CONFIG_USB_DWC2_HOST || CONFIG_USB_DWC2_DUAL_ROLE */
+
+#if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
+	/* Gadget structures */
+	struct usb_gadget_driver *driver;
+	struct phy *phy;
+	struct usb_phy *uphy;
+	struct s3c_hsotg_plat *plat;
+
+	int	irq;
+	struct clk *clk;
+
+	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
+
+	u32 phyif;
+	int fifo_mem;
+	unsigned int dedicated_fifos:1;
+	unsigned char num_of_eps;
+	u32 fifo_map;
+
+	struct dentry *debug_root;
+	struct dentry *debug_file;
+	struct dentry *debug_fifo;
+
+	struct usb_request *ep0_reply;
+	struct usb_request *ctrl_req;
+	u8 ep0_buff[8];
+	u8 ctrl_buff[8];
+
+	struct usb_gadget gadget;
+	unsigned int setup;
+	unsigned long last_rst;
+	struct s3c_hsotg_ep *eps;
+#endif /* CONFIG_USB_DWC2_PERIPHERAL || CONFIG_USB_DWC2_DUAL_ROLE */
 };
 
 /* Reasons for halting a host channel */
drivers/usb/dwc2/gadget.c +73 -72

571 lines not on this page · raw message

Signature · 1 line
2.0.3
#2 lore raw

Re: [PATCHv6 1/8] usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure

Quoted 245 lines · On Tue, Oct 28, 2014 at 06:25:42PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Adds the gadget data structure and appropriate data structure pointers to the common dwc2_hsotg data structure. To keep the driver data dereference code looking clean, the gadget variable declares are only available for peripheral and dual-role mode. This is needed so that the dwc2_hsotg data structure can be used by the hcd and gadget drivers.

Updates gadget.c to use the dwc2_hsotg data structure and gadget pointers that have been moved into the common dwc2_hsotg structure.

---
v5: Keep the changes to mininum and maintain hcd and gadget driver to build
and work separately. Use IS_ENABLED() instead of #if defined
v3: Updated with paulz's suggestion to avoid double pointers.
v2: Left the function parameter name as 'hsotg' and just changed its type.

---
 drivers/usb/dwc2/core.h   | 156 ++++++++++++++++++++++++----------------------
 drivers/usb/dwc2/gadget.c | 145 +++++++++++++++++++++---------------------
 2 files changed, 154 insertions(+), 147 deletions(-)
drivers/usb/dwc2/core.h +47 -75
index 55c90c5..96c283d 100644
@@ -84,7 +84,7 @@ static const char * const s3c_hsotg_supply_names[] = {
  */
 #define EP0_MPS_LIMIT   64
 
-struct s3c_hsotg;
+struct dwc2_hsotg;
 struct s3c_hsotg_req;
 
 /**
@@ -130,7 +130,7 @@ struct s3c_hsotg_req;
 struct s3c_hsotg_ep {
 	struct usb_ep           ep;
 	struct list_head        queue;
-	struct s3c_hsotg        *parent;
+	struct dwc2_hsotg       *parent;
 	struct s3c_hsotg_req    *req;
 	struct dentry           *debugfs;
 
@@ -155,67 +155,6 @@ struct s3c_hsotg_ep {
 };
 
 /**
- * struct s3c_hsotg - driver state.
- * @dev: The parent device supplied to the probe function
- * @driver: USB gadget driver
- * @phy: The otg phy transceiver structure for phy control.
- * @uphy: The otg phy transceiver structure for old USB phy control.
- * @plat: The platform specific configuration data. This can be removed once
- * all SoCs support usb transceiver.
- * @regs: The memory area mapped for accessing registers.
- * @irq: The IRQ number we are using
- * @supplies: Definition of USB power supplies
- * @phyif: PHY interface width
- * @dedicated_fifos: Set if the hardware has dedicated IN-EP fifos.
- * @num_of_eps: Number of available EPs (excluding EP0)
- * @debug_root: root directrory for debugfs.
- * @debug_file: main status file for debugfs.
- * @debug_fifo: FIFO status file for debugfs.
- * @ep0_reply: Request used for ep0 reply.
- * @ep0_buff: Buffer for EP0 reply data, if needed.
- * @ctrl_buff: Buffer for EP0 control requests.
- * @ctrl_req: Request for EP0 control packets.
- * @setup: NAK management for EP0 SETUP
- * @last_rst: Time of last reset
- * @eps: The endpoints being supplied to the gadget framework
- */
-struct s3c_hsotg {
-	struct device            *dev;
-	struct usb_gadget_driver *driver;
-	struct phy               *phy;
-	struct usb_phy           *uphy;
-	struct s3c_hsotg_plat    *plat;
-
-	spinlock_t              lock;
-
-	void __iomem            *regs;
-	int                     irq;
-	struct clk              *clk;
-
-	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
-
-	u32                     phyif;
-	int			fifo_mem;
-	unsigned int            dedicated_fifos:1;
-	unsigned char           num_of_eps;
-	u32			fifo_map;
-
-	struct dentry           *debug_root;
-	struct dentry           *debug_file;
-	struct dentry           *debug_fifo;
-
-	struct usb_request      *ep0_reply;
-	struct usb_request      *ctrl_req;
-	u8                      ep0_buff[8];
-	u8                      ctrl_buff[8];
-
-	struct usb_gadget       gadget;
-	unsigned int            setup;
-	unsigned long           last_rst;
-	struct s3c_hsotg_ep     *eps;
-};
-
-/**
  * struct s3c_hsotg_req - data transfer request
  * @req: The USB gadget request
  * @queue: The list of requests for the endpoint this is queued for.
@@ -229,6 +168,7 @@ struct s3c_hsotg_req {
 	unsigned char           mapped;
 };
 
+#if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 #define call_gadget(_hs, _entry) \
 do { \
 	if ((_hs)->gadget.speed != USB_SPEED_UNKNOWN && \
@@ -238,6 +178,9 @@ do { \
 		spin_lock(&_hs->lock); \
 	} \
 } while (0)
+#else
+#define call_gadget(_hs, _entry)	do {} while (0)
+#endif
 
 struct dwc2_hsotg;
 struct dwc2_host_chan;
@@ -495,11 +438,13 @@ struct dwc2_hw_params {
  * struct dwc2_hsotg - Holds the state of the driver, including the non-periodic
  * and periodic schedules
  *
+ * These are common for both host and peripheral modes:
+ *
  * @dev:                The struct device pointer
  * @regs:		Pointer to controller regs
- * @core_params:        Parameters that define how the core should be configured
  * @hw_params:          Parameters that were autodetected from the
  *                      hardware registers
+ * @core_params:	Parameters that define how the core should be configured
  * @op_state:           The operational State, during transitions (a_host=>
  *                      a_peripheral and b_device=>b_host) this may not match
  *                      the core, but allows the software to determine
@@ -508,6 +453,8 @@ struct dwc2_hw_params {
  *                      - USB_DR_MODE_PERIPHERAL
  *                      - USB_DR_MODE_HOST
  *                      - USB_DR_MODE_OTG
+ * @lock:		Spinlock that protects all the driver data structures
+ * @priv:		Stores a pointer to the struct usb_hcd
  * @queuing_high_bandwidth: True if multiple packets of a high-bandwidth
  *                      transfer are in process of being queued
  * @srp_success:        Stores status of SRP request in the case of a FS PHY
@@ -517,6 +464,9 @@ struct dwc2_hw_params {
  *                      interrupt
  * @wkp_timer:          Timer object for handling Wakeup Detected interrupt
  * @lx_state:           Lx state of connected device
+ *
+ * These are for host mode:
+ *
  * @flags:              Flags for handling root port state changes
  * @non_periodic_sched_inactive: Inactive QHs in the non-periodic schedule.
  *                      Transfers associated with these QHs are not currently
@@ -585,11 +535,31 @@ struct dwc2_hw_params {
  * @status_buf_dma:     DMA address for status_buf
  * @start_work:         Delayed work for handling host A-cable connection
  * @reset_work:         Delayed work for handling a port reset
- * @lock:               Spinlock that protects all the driver data structures
- * @priv:               Stores a pointer to the struct usb_hcd
  * @otg_port:           OTG port number
  * @frame_list:         Frame list
  * @frame_list_dma:     Frame list DMA address
+ *
+ * These are for peripheral mode:
+ *
+ * @driver:             USB gadget driver
+ * @phy:                The otg phy transceiver structure for phy control.
+ * @uphy:               The otg phy transceiver structure for old USB phy control.
+ * @plat:               The platform specific configuration data. This can be removed once
+ *                      all SoCs support usb transceiver.
+ * @supplies:           Definition of USB power supplies
+ * @phyif:              PHY interface width
+ * @dedicated_fifos:    Set if the hardware has dedicated IN-EP fifos.
+ * @num_of_eps:         Number of available EPs (excluding EP0)
+ * @debug_root:         Root directrory for debugfs.
+ * @debug_file:         Main status file for debugfs.
+ * @debug_fifo:         FIFO status file for debugfs.
+ * @ep0_reply:          Request used for ep0 reply.
+ * @ep0_buff:           Buffer for EP0 reply data, if needed.
+ * @ctrl_buff:          Buffer for EP0 control requests.
+ * @ctrl_req:           Request for EP0 control packets.
+ * @setup:              NAK management for EP0 SETUP
+ * @last_rst:           Time of last reset
+ * @eps:                The endpoints being supplied to the gadget framework
  */
 struct dwc2_hsotg {
 	struct device *dev;
@@ -601,6 +571,9 @@ struct dwc2_hsotg {
 	enum usb_otg_state op_state;
 	enum usb_dr_mode dr_mode;
 
+	spinlock_t lock;
+	void *priv;
+
 	unsigned int queuing_high_bandwidth:1;
 	unsigned int srp_success:1;
 
@@ -609,6 +582,14 @@ struct dwc2_hsotg {
 	struct timer_list wkp_timer;
 	enum dwc2_lx_state lx_state;
 
+	/* DWC OTG HW Release versions */
+#define DWC2_CORE_REV_2_71a	0x4f54271a
+#define DWC2_CORE_REV_2_90a	0x4f54290a
+#define DWC2_CORE_REV_2_92a	0x4f54292a
+#define DWC2_CORE_REV_2_94a	0x4f54294a
+#define DWC2_CORE_REV_3_00a	0x4f54300a
+
+#if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 	union dwc2_hcd_internal_flags {
 		u32 d32;
 		struct {
@@ -655,19 +636,10 @@ struct dwc2_hsotg {
 
 	struct delayed_work start_work;
 	struct delayed_work reset_work;
-	spinlock_t lock;
-	void *priv;
 	u8 otg_port;
 	u32 *frame_list;
 	dma_addr_t frame_list_dma;
 
-	/* DWC OTG HW Release versions */
-#define DWC2_CORE_REV_2_71a	0x4f54271a
-#define DWC2_CORE_REV_2_90a	0x4f54290a
-#define DWC2_CORE_REV_2_92a	0x4f54292a
-#define DWC2_CORE_REV_2_94a	0x4f54294a
-#define DWC2_CORE_REV_3_00a	0x4f54300a
-
 #ifdef DEBUG
 	u32 frrem_samples;
 	u64 frrem_accum;

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 666

Quoted 9 lines
drivers/usb/dwc2/core.h +6 -0
@@ -686,6 +658,40 @@ struct dwc2_hsotg {
 	u32 hfnum_other_samples_b;
 	u64 hfnum_other_frrem_accum_b;
 #endif
+#endif /* CONFIG_USB_DWC2_HOST || CONFIG_USB_DWC2_DUAL_ROLE */
+
+#if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
+	/* Gadget structures */
+	struct usb_gadget_driver *driver;
+	struct phy *phy;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 667

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct usb_phy *uphy;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 670

Quoted 3 lines
drivers/usb/dwc2/core.h +3 -0
+	struct s3c_hsotg_plat *plat;
+
+	int	irq;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 671

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct clk *clk;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 673

Quoted 2 lines
drivers/usb/dwc2/core.h +2 -0
+
+	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 681

Quoted 7 lines
drivers/usb/dwc2/core.h +7 -0
+	u32 phyif;
+	int fifo_mem;
+	unsigned int dedicated_fifos:1;
+	unsigned char num_of_eps;
+	u32 fifo_map;
+
+	struct dentry *debug_root;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 682

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct dentry *debug_file;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 683

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct dentry *debug_fifo;

this shouldn't be limited to gadget.

Signature · 1 line
balbi
#3 lore raw

Re: [PATCHv6 1/8] usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 204 lines · Felipe Balbi
Quoted 172 lines · On Tue, Oct 28, 2014 at 06:25:42PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Adds the gadget data structure and appropriate data structure pointers to the common dwc2_hsotg data structure. To keep the driver data dereference code looking clean, the gadget variable declares are only available for peripheral and dual-role mode. This is needed so that the dwc2_hsotg data structure can be used by the hcd and gadget drivers.

Updates gadget.c to use the dwc2_hsotg data structure and gadget pointers that have been moved into the common dwc2_hsotg structure.

the changes to mininum and maintain hcd and gadget driver to build and work separately. Use IS_ENABLED() instead of #if defined v3: Updated with paulz's suggestion to avoid double pointers. v2: Left the function parameter name as 'hsotg' and
just changed its type. --- drivers/usb/dwc2/core.h | 156
++++++++++++++++++++++++----------------------
drivers/usb/dwc2/gadget.c | 145 +++++++++++++++++++++--------------------- 2 files changed, 154 insertions(+), 147 deletions(-)

drivers/usb/dwc2/core.h +0 -0
index 55c90c5..96c283d 100644 --- a/drivers/usb/dwc2/core.h +++

b/drivers/usb/dwc2/core.h @@ -84,7 +84,7 @@ static const char *
const s3c_hsotg_supply_names[] = { */ #define EP0_MPS_LIMIT 64

-struct s3c_hsotg; +struct dwc2_hsotg; struct s3c_hsotg_req;

/** @@ -130,7 +130,7 @@ struct s3c_hsotg_req; struct s3c_hsotg_ep
{ struct usb_ep ep; struct list_head queue; -
struct s3c_hsotg *parent; + struct dwc2_hsotg
*parent; struct s3c_hsotg_req *req; struct dentry
*debugfs;

drivers/usb/dwc2/core.h +0 -0
@@ -155,67 +155,6 @@ struct s3c_hsotg_ep { };
 
 /** - * struct s3c_hsotg - driver state. - * @dev: The parent

device supplied to the probe function - * @driver: USB gadget driver - * @phy: The otg phy transceiver structure for phy control. - * @uphy: The otg phy transceiver structure for old USB phy control. - * @plat: The platform specific configuration data. This can be removed once - * all SoCs support usb transceiver. -
* @regs: The memory area mapped for accessing registers. - * @irq: The IRQ number we are using - * @supplies: Definition of USB power supplies - * @phyif: PHY interface width - * @dedicated_fifos: Set if the hardware has dedicated IN-EP fifos.
- * @num_of_eps: Number of available EPs (excluding EP0) - * @debug_root: root directrory for debugfs. - * @debug_file: main status file for debugfs. - * @debug_fifo: FIFO status file for debugfs. - * @ep0_reply: Request used for ep0 reply. - * @ep0_buff: Buffer for EP0 reply data, if needed. - * @ctrl_buff:
Buffer for EP0 control requests. - * @ctrl_req: Request for EP0 control packets. - * @setup: NAK management for EP0 SETUP - * @last_rst: Time of last reset - * @eps: The endpoints being supplied to the gadget framework - */ -struct s3c_hsotg { -
struct device *dev; - struct usb_gadget_driver
*driver; - struct phy *phy; - struct usb_phy
*uphy; - struct s3c_hsotg_plat *plat; - - spinlock_t
lock; - - void __iomem *regs; - int
irq; - struct clk *clk; - - struct
regulator_bulk_data
supplies[ARRAY_SIZE(s3c_hsotg_supply_names)]; - - u32
phyif; - int fifo_mem; - unsigned int
dedicated_fifos:1; - unsigned char num_of_eps; - u32
fifo_map; - - struct dentry *debug_root; - struct
dentry *debug_file; - struct dentry
*debug_fifo; - - struct usb_request *ep0_reply; - struct
usb_request *ctrl_req; - u8
ep0_buff[8]; - u8 ctrl_buff[8]; - - struct
usb_gadget gadget; - unsigned int setup; -
unsigned long last_rst; - struct s3c_hsotg_ep
*eps; -}; - -/** * struct s3c_hsotg_req - data transfer request * @req: The USB gadget request * @queue: The list of requests for the endpoint this is queued for. @@ -229,6 +168,7 @@ struct
s3c_hsotg_req { unsigned char mapped; };

+#if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE) #define call_gadget(_hs, _entry) \ do { \ if ((_hs)->gadget.speed != USB_SPEED_UNKNOWN && \ @@ -238,6 +178,9 @@ do { \ spin_lock(&_hs->lock); \ } \ } while
(0) +#else +#define call_gadget(_hs, _entry) do {} while (0)
+#endif

struct dwc2_hsotg; struct dwc2_host_chan; @@ -495,11 +438,13 @@ struct dwc2_hw_params { * struct dwc2_hsotg - Holds the state of the driver, including the non-periodic * and periodic schedules
* + * These are common for both host and peripheral modes: + * *
@dev: The struct device pointer * @regs: Pointer
to controller regs - * @core_params: Parameters that
define how the core should be configured * @hw_params:
Parameters that were autodetected from the *
hardware registers + * @core_params: Parameters that define how
the core should be configured * @op_state: The
operational State, during transitions (a_host=> * a_peripheral and b_device=>b_host) this may not match * the core, but allows the software to determine @@ -508,6 +453,8
@@ struct dwc2_hw_params { * -
USB_DR_MODE_PERIPHERAL * - USB_DR_MODE_HOST
* - USB_DR_MODE_OTG + * @lock: Spinlock
that protects all the driver data structures + * @priv: Stores a
pointer to the struct usb_hcd * @queuing_high_bandwidth: True if multiple packets of a high-bandwidth *
transfer are in process of being queued * @srp_success:
Stores status of SRP request in the case of a FS PHY @@ -517,6 +464,9 @@ struct dwc2_hw_params { *
interrupt * @wkp_timer: Timer object for handling Wakeup
Detected interrupt * @lx_state: Lx state of connected
device + * + * These are for host mode: + * * @flags:
Flags for handling root port state changes * @non_periodic_sched_inactive: Inactive QHs in the non-periodic
schedule. * Transfers associated with these
QHs are not currently @@ -585,11 +535,31 @@ struct dwc2_hw_params
{ * @status_buf_dma: DMA address for status_buf *
@start_work: Delayed work for handling host A-cable
connection * @reset_work: Delayed work for handling a
port reset - * @lock: Spinlock that protects all
the driver data structures - * @priv: Stores a
pointer to the struct usb_hcd * @otg_port: OTG port
number * @frame_list: Frame list * @frame_list_dma:
Frame list DMA address + * + * These are for peripheral mode: +
* + * @driver: USB gadget driver + * @phy:
The otg phy transceiver structure for phy control. + * @uphy:
The otg phy transceiver structure for old USB phy control. + *
@plat: The platform specific configuration data.
This can be removed once + * all SoCs
support usb transceiver. + * @supplies: Definition of
USB power supplies + * @phyif: PHY interface width +
* @dedicated_fifos: Set if the hardware has dedicated IN-EP
fifos. + * @num_of_eps: Number of available EPs
(excluding EP0) + * @debug_root: Root directrory for
debugfs. + * @debug_file: Main status file for debugfs. +
* @debug_fifo: FIFO status file for debugfs. + *
@ep0_reply: Request used for ep0 reply. + * @ep0_buff:
Buffer for EP0 reply data, if needed. + * @ctrl_buff:
Buffer for EP0 control requests. + * @ctrl_req: Request
for EP0 control packets. + * @setup: NAK management
for EP0 SETUP + * @last_rst: Time of last reset + *
@eps: The endpoints being supplied to the gadget
framework */ struct dwc2_hsotg { struct device *dev; @@ -601,6 +571,9 @@ struct dwc2_hsotg { enum usb_otg_state op_state; enum usb_dr_mode dr_mode;

drivers/usb/dwc2/core.h +1 -0
+	spinlock_t lock; +	void *priv; + unsigned int

queuing_high_bandwidth:1; unsigned int srp_success:1;

drivers/usb/dwc2/core.h +1 -0
@@ -609,6 +582,14 @@ struct dwc2_hsotg { struct timer_list
 wkp_timer; enum dwc2_lx_state lx_state;
 
+	/* DWC OTG HW Release versions */ +#define DWC2_CORE_REV_2_71a

0x4f54271a +#define DWC2_CORE_REV_2_90a 0x4f54290a +#define
DWC2_CORE_REV_2_92a 0x4f54292a +#define DWC2_CORE_REV_2_94a
0x4f54294a +#define DWC2_CORE_REV_3_00a 0x4f54300a + +#if
IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE) union dwc2_hcd_internal_flags { u32 d32; struct { @@ -655,19 +636,10 @@ struct dwc2_hsotg {

struct delayed_work start_work; struct delayed_work reset_work; -
spinlock_t lock; - void *priv; u8 otg_port; u32 *frame_list;
dma_addr_t frame_list_dma;

drivers/usb/dwc2/core.h +0 -1
-	/* DWC OTG HW Release versions */ -#define DWC2_CORE_REV_2_71a

0x4f54271a -#define DWC2_CORE_REV_2_90a 0x4f54290a -#define
DWC2_CORE_REV_2_92a 0x4f54292a -#define DWC2_CORE_REV_2_94a
0x4f54294a -#define DWC2_CORE_REV_3_00a 0x4f54300a - #ifdef
DEBUG u32 frrem_samples; u64 frrem_accum; @@ -686,6 +658,40 @@ struct dwc2_hsotg { u32 hfnum_other_samples_b; u64 hfnum_other_frrem_accum_b; #endif +#endif /* CONFIG_USB_DWC2_HOST

Quoted 4 lines
|| CONFIG_USB_DWC2_DUAL_ROLE */ + +#if
IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) ||

IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE) + /* Gadget structures */ +
struct usb_gadget_driver *driver; + struct phy *phy;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 667

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct usb_phy *uphy;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct s3c_hsotg_plat *plat; + +	int	irq;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 671

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct clk *clk;

this shouldn't be limited to gadget.

Quoted 2 lines

+ + struct regulator_bulk_data
supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];

this shouldn't be limited to gadget.

Quoted 3 lines
drivers/usb/dwc2/core.h +1 -0
+	u32 phyif; +	int fifo_mem; +	unsigned int dedicated_fifos:1; +

unsigned char num_of_eps; + u32 fifo_map; + + struct dentry
*debug_root;

this shouldn't be limited to gadget.

Quoted 8 lines

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 682

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct dentry *debug_file;

this shouldn't be limited to gadget.

drivers/usb/dwc2/core.h struct dwc2_hsotg { · line 683

Quoted 1 line
drivers/usb/dwc2/core.h +1 -0
+	struct dentry *debug_fifo;

this shouldn't be limited to gadget.

All fixed..

thanks,
Dinh
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iQIcBAEBAgAGBQJUU6LrAAoJEBmUBAuBoyj024wP/j7x+PW8rV8Aqzhkg4d1zSKE BE7/Axir8uPVTZfIOFBouHz6h0APztgW+by7DZdXCE8K6gUDAfzruxLRHNjctT6E KRj6A+i7qYnF/0+kc9b2InWYfS8NcOxths7r70RhJz7jlCphpaqlfhlicMSc0Ury Z/EeOYRpbwltDHGSBSkcPaw0yL4h/BQIRm1rRaUj+10cliSYwW5Re6mfiWVV+0pl WA56MM5xf7N9PO1Q+TPyStLyUTLyb3jiNwCOWfUCIkpHwJC9AIgbuBZXPMrSGOoL cMNoLHkgfY4HPFgzGlGxIgbHdkMec6A78FzSgz04P5+Uzh/MGKyrecordMbpQAPS DCGGWOsPrujAbWhgrus+cfxznIV3+4vhP+uC387RgLsYaJLRpAEpYN5aTABRU8u2 uEsbfYIhRGaGDQqTR0mzcKMXy9xoUMLNLkpiaqlGTs2Et0f1ZkElIn/dpefx417u Ocmfq7ab0cXgHnYNjdk9yE9JchgjUG5sn+INKbm/lLmXNyN3VHNZnBO42v1d78Gw sylySISc1aUJGXkuRJDPM4oC4jyPrerLB4BQbF0bQqaakz+U83+I+wZ7e9IOldO9 bPk5I7VZO8F8Ve0oP1WvuVZGeCnhrTd2isfx/orQ4J3uQFBD9hddGyaroDPtRIkA zoNVOSXVySi9EiPYT12u
=382E
-----END PGP SIGNATURE-----

#4 lore raw

Re: [PATCHv6 1/8] usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure

In reply to [email protected]

Hi Dinh

Quoted 10 lines · On 10/29/2014 07:25 AM, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Adds the gadget data structure and appropriate data structure pointers to the common dwc2_hsotg data structure. To keep the driver data dereference code looking clean, the gadget variable declares are only available for peripheral and dual-role mode. This is needed so that the dwc2_hsotg data structure can be used by the hcd and gadget drivers.

Updates gadget.c to use the dwc2_hsotg data structure and gadget pointers that have been moved into the common dwc2_hsotg structure.

Are we still need so much s3c prefix? Can we replace all the s3c into dwc2?

- Kever

#5 lore raw

Re: [PATCHv6 1/8] usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure

Quoted 10 lines · Kever Yang

Hi Dinh

Quoted 8 lines · On 10/29/2014 07:25 AM, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Adds the gadget data structure and appropriate data structure pointers to the common dwc2_hsotg data structure. To keep the driver data dereference code looking clean, the gadget variable declares are only available for peripheral and dual-role mode. This is needed so that the dwc2_hsotg data structure can be used by the hcd and gadget drivers.

Quoted 3 lines
Quoted 2 lines

Updates gadget.c to use the dwc2_hsotg data structure and gadget pointers that have been moved into the common dwc2_hsotg structure.

Are we still need so much s3c prefix? Can we replace all the s3c into dwc2?

separate patch, one which you can send if it bothers you so much.

Signature · 1 line
balbi
#6 lore raw

PATCH v6 2/8 usb: dwc2: Move gadget probe function into platform code

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

This patch will aggregate the probing of gadget/hcd driver into platform.c. The gadget probe funtion is converted into gadget_init that is now only responsible for gadget only initialization. All the gadget resources is now handled by platform.c

Since the host workqueue will not get initialized if the driver is configured for peripheral mode only. Thus we need to check for wq_otg before calling queue_work().

Also, we move spin_lock_init to common location for both host and gadget that is either in platform.c or pci.c.

We also ove suspend/resume code to common platform code, and update it to use the new PM API (struct dev_pm_ops).

Lastly, move the "samsung,s3c6400-hsotg" binding into dwc2_of_match_table.

---
v5: Reworked by squashing the following commits into this one:
* [PATCHv4 02/12] usb: dwc2: move "samsung,s3c6400-hsotg" into common platform
* [PATCHv4 03/12] usb: dwc2: Update the gadget driver to use common dwc2_hsotg structure
* [PATCHv4 09/12] usb: dwc2: initialize the spin_lock for both host and gadget
* [PATCHv4 10/12] usb: dwc2: Add suspend/resume for gadget
* [PATCHv4 11/12] usb: dwc2: check that the host work queue is valid Also use IS_ENABLED instead of #if defined

---
 drivers/usb/dwc2/core.h      | 34 ++++++++++++++-
 drivers/usb/dwc2/core_intr.c |  8 ++--
 drivers/usb/dwc2/gadget.c    | 99 ++++++++++----------------------------------
 drivers/usb/dwc2/hcd.c       |  1 -
 drivers/usb/dwc2/hcd.h       | 10 -----
 drivers/usb/dwc2/pci.c       |  1 +
 drivers/usb/dwc2/platform.c  | 32 ++++++++++++++
 7 files changed, 92 insertions(+), 93 deletions(-)
drivers/usb/dwc2/core.h +33 -1
index 96c283d..de2b194 100644
@@ -667,7 +667,6 @@ struct dwc2_hsotg {
 	struct usb_phy *uphy;
 	struct s3c_hsotg_plat *plat;
 
-	int	irq;
 	struct clk *clk;
 
 	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
@@ -961,4 +960,37 @@ extern void dwc2_dump_global_registers(struct dwc2_hsotg *hsotg);
  */
 extern u16 dwc2_get_otg_version(struct dwc2_hsotg *hsotg);
 
+/* Gadget defines */
+#if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
+extern int s3c_hsotg_remove(struct dwc2_hsotg *hsotg);
+extern int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2);
+extern int s3c_hsotg_resume(struct dwc2_hsotg *dwc2);
+extern int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq);
+#else
+static inline int s3c_hsotg_remove(struct dwc2_hsotg *dwc2)
+{ return 0; }
+static inline int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2)
+{ return 0; }
+static inline int s3c_hsotg_resume(struct dwc2_hsotg *dwc2)
+{ return 0; }
+static inline int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq)
+{ return 0; }
+#endif
+
+#if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
+extern int dwc2_hcd_get_frame_number(struct dwc2_hsotg *hsotg);
+extern void dwc2_hcd_disconnect(struct dwc2_hsotg *hsotg);
+extern void dwc2_hcd_start(struct dwc2_hsotg *hsotg);
+#else
+static inline void dwc2_set_all_params(struct dwc2_core_params *params, int value) {}
+static inline int dwc2_hcd_get_frame_number(struct dwc2_hsotg *hsotg)
+{ return 0; }
+static inline void dwc2_hcd_disconnect(struct dwc2_hsotg *hsotg) {}
+static inline void dwc2_hcd_start(struct dwc2_hsotg *hsotg) {}
+static inline void dwc2_hcd_remove(struct dwc2_hsotg *hsotg) {}
+static inline int dwc2_hcd_init(struct dwc2_hsotg *hsotg, int irq,
+				const struct dwc2_core_params *params)
+{ return 0; }
+#endif
+
 #endif /* __DWC2_CORE_H__ */
drivers/usb/dwc2/core_intr.c +5 -3
index c93918b..b176c2f 100644
@@ -287,9 +287,11 @@ static void dwc2_handle_conn_id_status_change_intr(struct dwc2_hsotg *hsotg)
 	 * Release lock before scheduling workq as it holds spinlock during
 	 * scheduling.
 	 */
-	spin_unlock(&hsotg->lock);
-	queue_work(hsotg->wq_otg, &hsotg->wf_otg);
-	spin_lock(&hsotg->lock);
+	if (hsotg->wq_otg) {
+		spin_unlock(&hsotg->lock);
+		queue_work(hsotg->wq_otg, &hsotg->wf_otg);
+		spin_lock(&hsotg->lock);
+	}
 
 	/* Clear interrupt */
 	writel(GINTSTS_CONIDSTSCHNG, hsotg->regs + GINTSTS);
drivers/usb/dwc2/gadget.c +21 -78

188 lines not on this page · raw message

drivers/usb/dwc2/hcd.c +0 -1
index 0a0e6f0..4a3cce0 100644
@@ -2839,7 +2839,6 @@ int dwc2_hcd_init(struct dwc2_hsotg *hsotg, int irq,
 
 	hcd->has_tt = 1;
 
-	spin_lock_init(&hsotg->lock);
 	((struct wrapper_priv_data *) &hcd->hcd_priv)->hsotg = hsotg;
 	hsotg->priv = hcd;
 
drivers/usb/dwc2/hcd.h +0 -10
index a12bb15..e69a843 100644
@@ -668,9 +668,6 @@ extern irqreturn_t dwc2_handle_hcd_intr(struct dwc2_hsotg *hsotg);
  */
 extern void dwc2_hcd_stop(struct dwc2_hsotg *hsotg);
 
-extern void dwc2_hcd_start(struct dwc2_hsotg *hsotg);
-extern void dwc2_hcd_disconnect(struct dwc2_hsotg *hsotg);
-
 /**
  * dwc2_hcd_is_b_host() - Returns 1 if core currently is acting as B host,
  * and 0 otherwise
@@ -680,13 +677,6 @@ extern void dwc2_hcd_disconnect(struct dwc2_hsotg *hsotg);
 extern int dwc2_hcd_is_b_host(struct dwc2_hsotg *hsotg);
 
 /**
- * dwc2_hcd_get_frame_number() - Returns current frame number
- *
- * @hsotg: The DWC2 HCD
- */
-extern int dwc2_hcd_get_frame_number(struct dwc2_hsotg *hsotg);
-
-/**
  * dwc2_hcd_dump_state() - Dumps hsotg state
  *
  * @hsotg: The DWC2 HCD
drivers/usb/dwc2/pci.c +1 -0
index c291fca..6d33ecf 100644
@@ -141,6 +141,7 @@ static int dwc2_driver_probe(struct pci_dev *dev,
 
 	pci_set_master(dev);
 
+	spin_lock_init(&hsotg->lock);
 	retval = dwc2_hcd_init(hsotg, dev->irq, &dwc2_module_params);
 	if (retval) {
 		pci_disable_device(dev);
drivers/usb/dwc2/platform.c +32 -0
index 121dbda..5783ed0 100644
@@ -121,6 +121,7 @@ static int dwc2_driver_remove(struct platform_device *dev)
 	struct dwc2_hsotg *hsotg = platform_get_drvdata(dev);
 
 	dwc2_hcd_remove(hsotg);
+	s3c_hsotg_remove(hsotg);
 
 	return 0;
 }
@@ -129,6 +130,7 @@ static const struct of_device_id dwc2_of_match_table[] = {
 	{ .compatible = "brcm,bcm2835-usb", .data = &params_bcm2835 },
 	{ .compatible = "rockchip,rk3066-usb", .data = &params_rk3066 },
 	{ .compatible = "snps,dwc2", .data = NULL },
+	{ .compatible = "samsung,s3c6400-hsotg", .data = NULL},
 	{},
 };
 MODULE_DEVICE_TABLE(of, dwc2_of_match_table);
@@ -204,6 +206,10 @@ static int dwc2_driver_probe(struct platform_device *dev)
 
 	hsotg->dr_mode = of_usb_get_dr_mode(dev->dev.of_node);
 
+	spin_lock_init(&hsotg->lock);
+	retval = dwc2_gadget_init(hsotg, irq);
+	if (retval)
+		return retval;
 	retval = dwc2_hcd_init(hsotg, irq, params);
 	if (retval)
 		return retval;
@@ -213,10 +219,36 @@ static int dwc2_driver_probe(struct platform_device *dev)
 	return retval;
 }
 
+static int dwc2_suspend(struct device *dev)
+{
+	struct dwc2_hsotg *dwc2 = dev_get_drvdata(dev);
+	int ret = 0;
+
+	if (dwc2_is_device_mode(dwc2))
+		ret = s3c_hsotg_suspend(dwc2);
+	return ret;
+}
+
+static int dwc2_resume(struct device *dev)
+{
+	struct dwc2_hsotg *dwc2 = dev_get_drvdata(dev);
+	int ret = 0;
+
+	if (dwc2_is_device_mode(dwc2))
+		ret = s3c_hsotg_resume(dwc2);
+
+	return ret;
+}
+
+static const struct dev_pm_ops dwc2_dev_pm_ops = {
+	SET_SYSTEM_SLEEP_PM_OPS(dwc2_suspend, dwc2_resume)
+};
+
 static struct platform_driver dwc2_platform_driver = {
 	.driver = {
 		.name = dwc2_driver_name,
 		.of_match_table = dwc2_of_match_table,
+		.pm = &dwc2_dev_pm_ops,
 	},
 	.probe = dwc2_driver_probe,
 	.remove = dwc2_driver_remove,
Signature · 1 line
2.0.3
#7 lore raw

Re: [PATCHv6 2/8] usb: dwc2: Move gadget probe function into platform code

Quoted 5 lines · On Tue, Oct 28, 2014 at 06:25:43PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

This patch will aggregate the probing of gadget/hcd driver into platform.c. The gadget probe funtion is converted into gadget_init that is now only responsible for gadget only initialization. All the gadget resources is now

s/resources is/resources are

Quoted 5 lines

handled by platform.c

Since the host workqueue will not get initialized if the driver is configured for peripheral mode only. Thus we need to check for wq_otg before calling queue_work().

this period character in the middle of the sentence doesn't make sense, perhaps a comma is what you want ? The sentence can be improved too:

"Since the host workqueue will not get initialized if the drier is configured for peripheral mode only, we add a check for wq_otg before calling queue_work()."

Quoted 5 lines

Also, we move spin_lock_init to common location for both host and gadget that is either in platform.c or pci.c.

We also ove suspend/resume code to common platform code, and update it to use the new PM API (struct dev_pm_ops).

updating to dev_pm_ops should really be a separate patch.

Signature · 1 line
balbi
#8 lore raw

Re: [PATCHv6 2/8] usb: dwc2: Move gadget probe function into platform code

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 25 lines · Felipe Balbi
Quoted 6 lines · On Tue, Oct 28, 2014 at 06:25:43PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

This patch will aggregate the probing of gadget/hcd driver into platform.c. The gadget probe funtion is converted into gadget_init that is now only responsible for gadget only initialization. All the gadget resources is now

s/resources is/resources are

Quoted 5 lines

handled by platform.c

Since the host workqueue will not get initialized if the driver is configured for peripheral mode only. Thus we need to check for wq_otg before calling queue_work().

this period character in the middle of the sentence doesn't make sense, perhaps a comma is what you want ? The sentence can be improved too:

"Since the host workqueue will not get initialized if the drier is configured for peripheral mode only, we add a check for wq_otg before calling queue_work()."

Quoted 8 lines
Quoted 5 lines

Also, we move spin_lock_init to common location for both host and gadget that is either in platform.c or pci.c.

We also ove suspend/resume code to common platform code, and update it to use the new PM API (struct dev_pm_ops).

updating to dev_pm_ops should really be a separate patch.

All fixed...

Thanks,
Dinh
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iQIcBAEBAgAGBQJUU6PAAAoJEBmUBAuBoyj0KJUP/RWpP9YkctpmZQgrb2xq0JaG eLCpOnKVYaTBCAhydJOKBBx0odYyxORlxTWbJCbZPvPhrZ7NDTINqEVWJOqYmtqe Xl8aEPYm+ckcfTliONK3yIP/MA49he/qbGOmqXFwFMbcAvKFN/kpq05cjzZ6T8bO hILDO2Yy2HVQHonq3uKppmTB9DSnCNfCR0Cuum2fpOzmVLY/X46EM5UD2e+XmVgo dsINAGe40FwFIoLZzAradn12MrmCjOM6MPpkkGZ3SMGjKXY7yQX38c9WbcCvsm6g y1/5fDqdRlsq+Weoo9N3H03LUX8MlGCpZ7dgLrABXMEjdySj3eVTj4bHgPxRjRY/ +SfI2YR8RyEcHj2UTqgMUp4JmRl2CNiB9fsvZHMQQm2MTeFkcUpZnSUXWHB9Vtwv s1I5nMOSoT2NDBg6QS9a1T1s1gdSywOnDBd+/SK7mf4QkQumvf2Nvu6avQB1Rxxm sqpGWztuhuycg332CXc9W0EBXcg5t8SRyk9SFgksH3MezS85gAQxZtXyv44NuCUM ciTru8tKh3ncszws2NmX9yyQgC4fM3kHDytOwDXFt23f3xHyjg+2m3O1TLErAwig yqj2MfYTfqfYVrs/hOvOpEbUgUSy5MQY/EOMIsI6aGfB2Nba72wrYQP79ugb84u9 Ftd1XH47lvXX/R8VOV2x
=ZAkf
-----END PGP SIGNATURE-----

#9 lore raw

PATCH v6 3/8 usb: dwc2: Initialize the USB core for peripheral mode

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

Initialize the USB driver to peripheral mode when a B-Device connector is attached.

---
v5: move the export of s3c_hsotg_core_init into this patch

---
 drivers/usb/dwc2/core.h   | 2 ++
 drivers/usb/dwc2/gadget.c | 2 +-
 drivers/usb/dwc2/hcd.c    | 1 +
 3 files changed, 4 insertions(+), 1 deletion(-)
drivers/usb/dwc2/core.h +2 -0
index de2b194..80d29c7 100644
@@ -966,6 +966,7 @@ extern int s3c_hsotg_remove(struct dwc2_hsotg *hsotg);
 extern int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2);
 extern int s3c_hsotg_resume(struct dwc2_hsotg *dwc2);
 extern int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq);
+extern void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2);
 #else
 static inline int s3c_hsotg_remove(struct dwc2_hsotg *dwc2)
 { return 0; }
@@ -975,6 +976,7 @@ static inline int s3c_hsotg_resume(struct dwc2_hsotg *dwc2)
 { return 0; }
 static inline int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq)
 { return 0; }
+static inline void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2) {}
 #endif
 
 #if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
drivers/usb/dwc2/gadget.c +1 -1
index 38ec1cc..19d1b03 100644
@@ -2125,7 +2125,7 @@ static int s3c_hsotg_corereset(struct dwc2_hsotg *hsotg)
  *
  * Issue a soft reset to the core, and await the core finishing it.
  */
-static void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg)
+void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg)
 {
 	s3c_hsotg_corereset(hsotg);
 
drivers/usb/dwc2/hcd.c +1 -0
index 4a3cce0..44c609f 100644
@@ -1371,6 +1371,7 @@ static void dwc2_conn_id_status_change(struct work_struct *work)
 		hsotg->op_state = OTG_STATE_B_PERIPHERAL;
 		dwc2_core_init(hsotg, false, -1);
 		dwc2_enable_global_interrupts(hsotg);
+		s3c_hsotg_core_init(hsotg);
 	} else {
 		/* A-Device connector (Host Mode) */
 		dev_dbg(hsotg->dev, "connId A\n");
Signature · 1 line
2.0.3
#10 lore raw

PATCH v6 4/8 usb: dwc2: Update common interrupt handler to call gadget interrupt handler

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

Make dwc2_handle_common_intr call the gadget interrupt function when operating in peripheral mode. Remove the spinlock functions in s3c_hsotg_irq as dwc2_handle_common_intr() already has the spinlocks.

Move the registeration of the IRQ to common code for platform and PCI.

Remove duplicate interrupt conditions that was in gadget, as those are handled by dwc2 common interrupt handler.

---
v5: remove individual devm_request_irq from gadget and hcd, and place a
single devm_request_irq in platform and pci.
v2: Keep interrupt handler for host and peripheral modes separate

---
 drivers/usb/dwc2/core.c      | 10 --------
 drivers/usb/dwc2/core.h      |  3 +++
 drivers/usb/dwc2/core_intr.c |  3 +++
 drivers/usb/dwc2/gadget.c    | 57 ++------------------------------------------
 drivers/usb/dwc2/pci.c       |  6 +++++
 drivers/usb/dwc2/platform.c  |  9 +++++++
 6 files changed, 23 insertions(+), 65 deletions(-)
drivers/usb/dwc2/core.c +0 -10
index d926945..7605850b 100644
@@ -458,16 +458,6 @@ int dwc2_core_init(struct dwc2_hsotg *hsotg, bool select_phy, int irq)
 	/* Clear the SRP success bit for FS-I2c */
 	hsotg->srp_success = 0;
 
-	if (irq >= 0) {
-		dev_dbg(hsotg->dev, "registering common handler for irq%d\n",
-			irq);
-		retval = devm_request_irq(hsotg->dev, irq,
-					  dwc2_handle_common_intr, IRQF_SHARED,
-					  dev_name(hsotg->dev), hsotg);
-		if (retval)
-			return retval;
-	}
-
 	/* Enable common interrupts */
 	dwc2_enable_common_interrupts(hsotg);
 
drivers/usb/dwc2/core.h +3 -0
index 80d29c7..ec70862 100644
@@ -967,6 +967,7 @@ extern int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2);
 extern int s3c_hsotg_resume(struct dwc2_hsotg *dwc2);
 extern int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq);
 extern void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2);
+irqreturn_t s3c_hsotg_irq(int irq, void *pw);
 #else
 static inline int s3c_hsotg_remove(struct dwc2_hsotg *dwc2)
 { return 0; }
@@ -977,6 +978,8 @@ static inline int s3c_hsotg_resume(struct dwc2_hsotg *dwc2)
 static inline int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq)
 { return 0; }
 static inline void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2) {}
+static inline irqreturn_t s3c_hsotg_irq(int irq, void *pw)
+{ return IRQ_HANDLED; }
 #endif
 
 #if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
drivers/usb/dwc2/core_intr.c +3 -0
index b176c2f..b0c14e0 100644
@@ -474,6 +474,9 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 
 	spin_lock(&hsotg->lock);
 
+	if (dwc2_is_device_mode(hsotg))
+		retval = s3c_hsotg_irq(irq, dev);
+
 	gintsts = dwc2_read_common_intr(hsotg);
 	if (gintsts & ~GINTSTS_PRTINT)
 		retval = IRQ_HANDLED;
drivers/usb/dwc2/gadget.c +2 -55

101 lines not on this page · raw message

drivers/usb/dwc2/pci.c +6 -0
index 6d33ecf..a4e724b 100644
@@ -141,6 +141,12 @@ static int dwc2_driver_probe(struct pci_dev *dev,
 
 	pci_set_master(dev);
 
+	retval = devm_request_irq(hsotg->dev, dev->irq,
+				  dwc2_handle_common_intr, IRQF_SHARED,
+				  dev_name(hsotg->dev), hsotg);
+	if (retval)
+		return retval;
+
 	spin_lock_init(&hsotg->lock);
 	retval = dwc2_hcd_init(hsotg, dev->irq, &dwc2_module_params);
 	if (retval) {
drivers/usb/dwc2/platform.c +9 -0
index 5783ed0..72f32f7 100644
@@ -196,6 +196,15 @@ static int dwc2_driver_probe(struct platform_device *dev)
 		return irq;
 	}
 
+	dev_dbg(hsotg->dev, "registering common handler for irq%d\n",
+		irq);
+
+	retval = devm_request_irq(hsotg->dev, irq,
+				  dwc2_handle_common_intr, IRQF_SHARED,
+				  dev_name(hsotg->dev), hsotg);
+	if (retval)
+		return retval;
+
 	res = platform_get_resource(dev, IORESOURCE_MEM, 0);
 	hsotg->regs = devm_ioremap_resource(&dev->dev, res);
 	if (IS_ERR(hsotg->regs))
Signature · 1 line
2.0.3
#11 lore raw

Re: [PATCHv6 4/8] usb: dwc2: Update common interrupt handler to call gadget interrupt handler

Hi,

Quoted 87 lines · On Tue, Oct 28, 2014 at 06:25:45PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Make dwc2_handle_common_intr call the gadget interrupt function when operating in peripheral mode. Remove the spinlock functions in s3c_hsotg_irq as dwc2_handle_common_intr() already has the spinlocks.

Move the registeration of the IRQ to common code for platform and PCI.

Remove duplicate interrupt conditions that was in gadget, as those are handled by dwc2 common interrupt handler.

---
v5: remove individual devm_request_irq from gadget and hcd, and place a
single devm_request_irq in platform and pci.
v2: Keep interrupt handler for host and peripheral modes separate

---
 drivers/usb/dwc2/core.c      | 10 --------
 drivers/usb/dwc2/core.h      |  3 +++
 drivers/usb/dwc2/core_intr.c |  3 +++
 drivers/usb/dwc2/gadget.c    | 57 ++------------------------------------------
 drivers/usb/dwc2/pci.c       |  6 +++++
 drivers/usb/dwc2/platform.c  |  9 +++++++
 6 files changed, 23 insertions(+), 65 deletions(-)
drivers/usb/dwc2/core.c +0 -10
index d926945..7605850b 100644
@@ -458,16 +458,6 @@ int dwc2_core_init(struct dwc2_hsotg *hsotg, bool select_phy, int irq)
 	/* Clear the SRP success bit for FS-I2c */
 	hsotg->srp_success = 0;
 
-	if (irq >= 0) {
-		dev_dbg(hsotg->dev, "registering common handler for irq%d\n",
-			irq);
-		retval = devm_request_irq(hsotg->dev, irq,
-					  dwc2_handle_common_intr, IRQF_SHARED,
-					  dev_name(hsotg->dev), hsotg);
-		if (retval)
-			return retval;
-	}
-
 	/* Enable common interrupts */
 	dwc2_enable_common_interrupts(hsotg);
 
drivers/usb/dwc2/core.h +3 -0
index 80d29c7..ec70862 100644
@@ -967,6 +967,7 @@ extern int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2);
 extern int s3c_hsotg_resume(struct dwc2_hsotg *dwc2);
 extern int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq);
 extern void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2);
+irqreturn_t s3c_hsotg_irq(int irq, void *pw);
 #else
 static inline int s3c_hsotg_remove(struct dwc2_hsotg *dwc2)
 { return 0; }
@@ -977,6 +978,8 @@ static inline int s3c_hsotg_resume(struct dwc2_hsotg *dwc2)
 static inline int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq)
 { return 0; }
 static inline void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2) {}
+static inline irqreturn_t s3c_hsotg_irq(int irq, void *pw)
+{ return IRQ_HANDLED; }
 #endif
 
 #if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
drivers/usb/dwc2/core_intr.c +3 -0
index b176c2f..b0c14e0 100644
@@ -474,6 +474,9 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 
 	spin_lock(&hsotg->lock);
 
+	if (dwc2_is_device_mode(hsotg))
+		retval = s3c_hsotg_irq(irq, dev);
+
 	gintsts = dwc2_read_common_intr(hsotg);
 	if (gintsts & ~GINTSTS_PRTINT)
 		retval = IRQ_HANDLED;
drivers/usb/dwc2/gadget.c +0 -0
index 19d1b03..202f8cc 100644

drivers/usb/dwc2/gadget.c void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg) · line 2260

Quoted 5 lines
drivers/usb/dwc2/gadget.c +1 -1
@@ -2257,14 +2257,13 @@ void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg)
  * @irq: The IRQ number triggered
  * @pw: The pw value when registered the handler.
  */
-static irqreturn_t s3c_hsotg_irq(int irq, void *pw)
+irqreturn_t s3c_hsotg_irq(int irq, void *pw)

why ? It would've been a lot easier to just make the IRQ line shared.

Signature · 1 line
balbi
#12 lore raw

Re: [PATCHv6 4/8] usb: dwc2: Update common interrupt handler to call gadget interrupt handler

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 70 lines · Felipe Balbi

Hi,

Quoted 66 lines · On Tue, Oct 28, 2014 at 06:25:45PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Make dwc2_handle_common_intr call the gadget interrupt function when operating in peripheral mode. Remove the spinlock functions in s3c_hsotg_irq as dwc2_handle_common_intr() already has the spinlocks.

Move the registeration of the IRQ to common code for platform and PCI.

Remove duplicate interrupt conditions that was in gadget, as those are handled by dwc2 common interrupt handler.

individual devm_request_irq from gadget and hcd, and place a single devm_request_irq in platform and pci. v2: Keep interrupt handler for host and peripheral modes separate ---

drivers/usb/dwc2/core.c      | 10 -------- 
drivers/usb/dwc2/core.h      |  3 +++ 
drivers/usb/dwc2/core_intr.c |  3 +++ drivers/usb/dwc2/gadget.c
| 57 ++------------------------------------------ 
drivers/usb/dwc2/pci.c       |  6 +++++ 
drivers/usb/dwc2/platform.c  |  9 +++++++ 6 files changed, 23

insertions(+), 65 deletions(-)

drivers/usb/dwc2/core.c +0 -0
index d926945..7605850b 100644 --- a/drivers/usb/dwc2/core.c +++

b/drivers/usb/dwc2/core.c @@ -458,16 +458,6 @@ int dwc2_core_init(struct dwc2_hsotg *hsotg, bool select_phy, int irq) /* Clear the SRP success bit for FS-I2c */ hsotg->srp_success = 0;

drivers/usb/dwc2/core.c +0 -1
-	if (irq >= 0) { -		dev_dbg(hsotg->dev, "registering common

handler for irq%d\n", - irq); - retval =
devm_request_irq(hsotg->dev, irq, -
dwc2_handle_common_intr, IRQF_SHARED, -
dev_name(hsotg->dev), hsotg); - if (retval) - return retval; -
} - /* Enable common interrupts */
dwc2_enable_common_interrupts(hsotg);

drivers/usb/dwc2/core.h +0 -0
index 80d29c7..ec70862 100644 --- a/drivers/usb/dwc2/core.h +++

b/drivers/usb/dwc2/core.h @@ -967,6 +967,7 @@ extern int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2); extern int s3c_hsotg_resume(struct dwc2_hsotg *dwc2); extern int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq); extern void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2); +irqreturn_t s3c_hsotg_irq(int irq, void *pw); #else static inline int s3c_hsotg_remove(struct dwc2_hsotg *dwc2) { return 0; } @@ -977,6 +978,8 @@ static inline int s3c_hsotg_resume(struct dwc2_hsotg *dwc2) static inline int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq) { return 0; } static inline void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2) {} +static inline irqreturn_t s3c_hsotg_irq(int irq, void *pw) +{ return IRQ_HANDLED; } #endif

#if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE) diff --git a/drivers/usb/dwc2/core_intr.c b/drivers/usb/dwc2/core_intr.c index b176c2f..b0c14e0 100644 --- a/drivers/usb/dwc2/core_intr.c +++ b/drivers/usb/dwc2/core_intr.c @@ -474,6 +474,9 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)

spin_lock(&hsotg->lock);

Quoted 14 lines
Quoted 10 lines
drivers/usb/dwc2/core.h +1 -0
+	if (dwc2_is_device_mode(hsotg)) +		retval = s3c_hsotg_irq(irq,

dev); + gintsts = dwc2_read_common_intr(hsotg); if (gintsts & ~GINTSTS_PRTINT) retval = IRQ_HANDLED; diff --git a/drivers/usb/dwc2/gadget.c b/drivers/usb/dwc2/gadget.c index 19d1b03..202f8cc 100644 --- a/drivers/usb/dwc2/gadget.c +++ b/drivers/usb/dwc2/gadget.c @@ -2257,14 +2257,13 @@ void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg) * @irq: The IRQ number triggered * @pw: The pw value when registered the handler. */ -static irqreturn_t s3c_hsotg_irq(int irq, void *pw) +irqreturn_t s3c_hsotg_irq(int irq, void *pw)

why ? It would've been a lot easier to just make the IRQ line shared.

Reworked to use IRQF_SHARED...

Thanks,
Dinh
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iQIcBAEBAgAGBQJUU6P5AAoJEBmUBAuBoyj0vWwP/1Qq0Gt9U1MF6LHJU853+P+I jx6xVVanPSuK8yTWprh7M/endv+UkVwwt+2sGZ0MMzLDH1T7Kf6FdqhjlNqvFeRK qK5LUkpMKvMJKcw6tQNGsjB75X93GRkWLxuZEQ9BpGjfy1fct6Wic/kv66dUVCCI DevfZZFPGdSmiichVup0sO/cfTJ84R8xyQVuF77LvZ/Wbm606ypV4PTMEjKcFhmr 5s6BeOINm7p1NMuoDaWbmjjOHAfTwy06W3p81LMsWH5yfVCq4TgJSwMbdo8s/i/Y po/1Nu0x5lwwao1J2OZA7b30i1vbtztCzVPtFcO5quN9rOlsaXRNBGLf9jjHP1OO PzRSsh1EbAinc4Ohbnii/bdx2g7lWP2ktf5qqoeVocgZkrKA/k5w/QK1qL1z3Onl ZkylCf5VMd06eOkeGbdce8UFhbceNLw0ryqAVN0bsxiD0u78aH6KvhYhBo29p/1R VMlKPrsYVDp9xjsp24S/u7o4NvNx8VeIfaKA5JN0Hnzncy03ByaO/o5iMPLb231l /yCoVyt2A9q7LQs/HeyCZYOFAPpG8ZbTsZrucKB9rRWXizDIrpIfJdcH3mIp71ab fhhf8sE1kDMrqADLTd1UL96DKCJnU2N8qJLSll6uWNi95J6OogIkj+H4YFPXXWiM +wUVqVAyYzPKsCaMRNTF
=5+gL
-----END PGP SIGNATURE-----

#13 lore raw

RE: [PATCHv6 4/8] usb: dwc2: Update common interrupt handler to call gadget interrupt handler

In reply to Felipe Balbi
Quoted 90 lines

From: Felipe Balbi [mailto:[email protected]]
Sent: Thursday, October 30, 2014 7:01 AM

Quoted 86 lines · On Tue, Oct 28, 2014 at 06:25:45PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Make dwc2_handle_common_intr call the gadget interrupt function when operating in peripheral mode. Remove the spinlock functions in s3c_hsotg_irq as dwc2_handle_common_intr() already has the spinlocks.

Move the registeration of the IRQ to common code for platform and PCI.

Remove duplicate interrupt conditions that was in gadget, as those are handled by dwc2 common interrupt handler.

---
v5: remove individual devm_request_irq from gadget and hcd, and place a
single devm_request_irq in platform and pci.
v2: Keep interrupt handler for host and peripheral modes separate

---
 drivers/usb/dwc2/core.c      | 10 --------
 drivers/usb/dwc2/core.h      |  3 +++
 drivers/usb/dwc2/core_intr.c |  3 +++
 drivers/usb/dwc2/gadget.c    | 57 ++------------------------------------------
 drivers/usb/dwc2/pci.c       |  6 +++++
 drivers/usb/dwc2/platform.c  |  9 +++++++
 6 files changed, 23 insertions(+), 65 deletions(-)
drivers/usb/dwc2/core.c +0 -10
index d926945..7605850b 100644
@@ -458,16 +458,6 @@ int dwc2_core_init(struct dwc2_hsotg *hsotg, bool select_phy, int irq)
 	/* Clear the SRP success bit for FS-I2c */
 	hsotg->srp_success = 0;
 
-	if (irq >= 0) {
-		dev_dbg(hsotg->dev, "registering common handler for irq%d\n",
-			irq);
-		retval = devm_request_irq(hsotg->dev, irq,
-					  dwc2_handle_common_intr, IRQF_SHARED,
-					  dev_name(hsotg->dev), hsotg);
-		if (retval)
-			return retval;
-	}
-
 	/* Enable common interrupts */
 	dwc2_enable_common_interrupts(hsotg);
 
drivers/usb/dwc2/core.h +3 -0
index 80d29c7..ec70862 100644
@@ -967,6 +967,7 @@ extern int s3c_hsotg_suspend(struct dwc2_hsotg *dwc2);
 extern int s3c_hsotg_resume(struct dwc2_hsotg *dwc2);
 extern int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq);
 extern void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2);
+irqreturn_t s3c_hsotg_irq(int irq, void *pw);
 #else
 static inline int s3c_hsotg_remove(struct dwc2_hsotg *dwc2)
 { return 0; }
@@ -977,6 +978,8 @@ static inline int s3c_hsotg_resume(struct dwc2_hsotg *dwc2)
 static inline int dwc2_gadget_init(struct dwc2_hsotg *hsotg, int irq)
 { return 0; }
 static inline void s3c_hsotg_core_init(struct dwc2_hsotg *dwc2) {}
+static inline irqreturn_t s3c_hsotg_irq(int irq, void *pw)
+{ return IRQ_HANDLED; }
 #endif
 
 #if IS_ENABLED(CONFIG_USB_DWC2_HOST) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
drivers/usb/dwc2/core_intr.c +3 -0
index b176c2f..b0c14e0 100644
@@ -474,6 +474,9 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 
 	spin_lock(&hsotg->lock);
 
+	if (dwc2_is_device_mode(hsotg))
+		retval = s3c_hsotg_irq(irq, dev);
+
 	gintsts = dwc2_read_common_intr(hsotg);
 	if (gintsts & ~GINTSTS_PRTINT)
 		retval = IRQ_HANDLED;
drivers/usb/dwc2/gadget.c +0 -0
index 19d1b03..202f8cc 100644
Quoted 8 lines

drivers/usb/dwc2/gadget.c void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg) · line 2260

Quoted 6 lines
drivers/usb/dwc2/gadget.c +1 -1
@@ -2257,14 +2257,13 @@ void s3c_hsotg_core_init(struct dwc2_hsotg *hsotg)
  * @irq: The IRQ number triggered
  * @pw: The pw value when registered the handler.
  */
-static irqreturn_t s3c_hsotg_irq(int irq, void *pw)
+irqreturn_t s3c_hsotg_irq(int irq, void *pw)

why ? It would've been a lot easier to just make the IRQ line shared.

Actually, I would prefer to keep the common interrupt handler. Reason being, the HSOTG core can spew a *lot* of interrupts when in host mode, and the driver already struggles to keep up with them on some platforms. We see this problem especially on the Raspberry Pi.

So I think adding the overhead of an additional interrupt handling path is a bad idea. Unless the peripheral-mode interrupt handler can somehow be disabled when the controller is in host mode.

Maybe we could delay registering the gadget interrupt handler until the controller switches to peripheral mode, and unregister it when it switches back to host mode? But that seems pretty "odd" to me.

Felipe, what do you think?

Signature · 1 line
Paul
#14 lore raw

PATCH v6 5/8 usb: dwc2: Add call_gadget functions for perpheral mode interrupts

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

Update the dwc2 wakeup and suspend interrupt functions to use call_gadget when the IP is in peripheral mode.

---
 drivers/usb/dwc2/core_intr.c | 2 ++
 1 file changed, 2 insertions(+)
drivers/usb/dwc2/core_intr.c +2 -0
index b0c14e0..1240875 100644
@@ -339,6 +339,7 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
+		call_gadget(hsotg, resume);
 	} else {
 		if (hsotg->lx_state != DWC2_L1) {
 			u32 pcgcctl = readl(hsotg->regs + PCGCTL);
@@ -399,6 +400,7 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
+		call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");
Signature · 1 line
2.0.3
#15 lore raw

Re: [PATCHv6 5/8] usb: dwc2: Add call_gadget functions for perpheral mode interrupts

Quoted 4 lines · On Tue, Oct 28, 2014 at 06:25:46PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Update the dwc2 wakeup and suspend interrupt functions to use call_gadget when the IP is in peripheral mode.

it seems like you're actually fixing a bug here. Those calls weren't there before.

Signature · 1 line
balbi
#16 lore raw

Re: [PATCHv6 5/8] usb: dwc2: Add call_gadget functions for perpheral mode interrupts

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Quoted 10 lines · Felipe Balbi
Quoted 4 lines · On Tue, Oct 28, 2014 at 06:25:46PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Update the dwc2 wakeup and suspend interrupt functions to use call_gadget when the IP is in peripheral mode.

it seems like you're actually fixing a bug here. Those calls weren't there before.

This patch can be sent separately from this series.

Thanks,
Dinh
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.11 (GNU/Linux)

iQIcBAEBAgAGBQJUU6QzAAoJEBmUBAuBoyj0XpIQAKLQxsksyfEBAOWNwapO1Grx jMK9+QpqZuGwo5Gha905iAO1Q/XP5tHvkLBzGyeEpNhEqBWXTTAvu1i2EToSiNId JTtae/eJA5xfxoljULqL3Vuy1SfOt+EdBHoMdn/vQQfBp9kUiqfJNgZeUsDgvbBK 7xNfdocMQjqPPbseNGL3193aasOFyZdc9sI+ZR3XjeXjNPgzwJzfB9eMZH+zpRKU nVcWY5UBkw05UnvDNwc4e/+VNpBRj4NLBBLGPleUJ1KXyunUet5r4oCLYLasZaFn zk6VvovQMUDk9DhnK2d8w6oosmSA/eD8zKn3K3++LV/XGwb5rHP+zC5jpAkbjN/o 4ZudegACKNUiwvo0tfbgtI5no/VzlaArSM6n3WT6Fz5cDveibZbU7z78AqIqRkwF Ai4+bn510OCgfEEu2rVRXl21r/TAt9r3ujh1mPPUwJ+nWdoxjmm7+0MyKNSeB/Ab MmVZfQaELLGTOMHH0T0/yr6XX8Bc4LxGtbjVQRlmW7mYfafnzhE+aWAeLnyF4V9r cA0YE09jOQoXh8WU6k18SKbapF8HDaXSLBcTWnXBR93TQu4U05qx5iB8hEZ7S9NT z3uQdfgjk8LjhaYWTtSsT29E6oDbvho8h5TF48KiwN4Ud0wagqJq1W3vlGhrgKUd wy94/mOPpifATM0S4KUr
=qCZw
-----END PGP SIGNATURE-----

#17 lore raw

PATCH v6 6/8 usb: dwc2: gadget: Do not fail probe if there isn't a clock node

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

Since the dwc2 hcd driver is currently not looking for a clock node during init, we should not completely fail if there isn't a clock provided. For dual-role mode, we will only fail init for a non-clock node error. We then update the HCD to only call gadget funtions if there is a proper clock node.

---
v5: reworked to not access gadget functions from the hcd.

---
 drivers/usb/dwc2/core.h      |  3 +--
 drivers/usb/dwc2/core_intr.c |  9 ++++++---
 drivers/usb/dwc2/hcd.c       |  3 ++-
 drivers/usb/dwc2/platform.c  | 19 +++++++++++++++----
 4 files changed, 24 insertions(+), 10 deletions(-)
drivers/usb/dwc2/core.h +1 -2
index ec70862..48120c8 100644
@@ -660,6 +660,7 @@ struct dwc2_hsotg {
 #endif
 #endif /* CONFIG_USB_DWC2_HOST || CONFIG_USB_DWC2_DUAL_ROLE */
 
+	struct clk *clk;
 #if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 	/* Gadget structures */
 	struct usb_gadget_driver *driver;
@@ -667,8 +668,6 @@ struct dwc2_hsotg {
 	struct usb_phy *uphy;
 	struct s3c_hsotg_plat *plat;
 
-	struct clk *clk;
-
 	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
 
 	u32 phyif;
drivers/usb/dwc2/core_intr.c +6 -3
index 1240875..1608037 100644
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);
 	} else {
 		if (hsotg->lx_state != DWC2_L1) {
 			u32 pcgcctl = readl(hsotg->regs + PCGCTL);
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);
 
 	gintsts = dwc2_read_common_intr(hsotg);
 	if (gintsts & ~GINTSTS_PRTINT)
drivers/usb/dwc2/hcd.c +2 -1
index 44c609f..fa49c72 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct work_struct *work)
 		hsotg->op_state = OTG_STATE_B_PERIPHERAL;
 		dwc2_core_init(hsotg, false, -1);
 		dwc2_enable_global_interrupts(hsotg);
-		s3c_hsotg_core_init(hsotg);
+		if (!IS_ERR(hsotg->clk))
+			s3c_hsotg_core_init(hsotg);
 	} else {
 		/* A-Device connector (Host Mode) */
 		dev_dbg(hsotg->dev, "connId A\n");
drivers/usb/dwc2/platform.c +15 -4
index 72f32f7..77c8417 100644
@@ -217,8 +217,17 @@ static int dwc2_driver_probe(struct platform_device *dev)
 
 	spin_lock_init(&hsotg->lock);
 	retval = dwc2_gadget_init(hsotg, irq);
-	if (retval)
-		return retval;
+	if (retval) {
+		/*
+		 * We will not fail the driver initialization for dual-role
+		 * if no clock node is supplied. However, all gadget
+		 * functionality will be disabled if a clock node is not
+		 * provided. Host functionality will continue.
+		 * TO-DO: make clock node a requirement for the HCD.
+		 */
+		if (!IS_ERR(hsotg->clk))
+			return retval;
+	}
 	retval = dwc2_hcd_init(hsotg, irq, params);
 	if (retval)
 		return retval;
@@ -234,7 +243,8 @@ static int dwc2_suspend(struct device *dev)
 	int ret = 0;
 
 	if (dwc2_is_device_mode(dwc2))
-		ret = s3c_hsotg_suspend(dwc2);
+		if (!IS_ERR(dwc2->clk))
+			ret = s3c_hsotg_suspend(dwc2);
 	return ret;
 }
 
@@ -244,7 +254,8 @@ static int dwc2_resume(struct device *dev)
 	int ret = 0;
 
 	if (dwc2_is_device_mode(dwc2))
-		ret = s3c_hsotg_resume(dwc2);
+		if (!IS_ERR(dwc2->clk))
+			ret = s3c_hsotg_resume(dwc2);
 
 	return ret;
 }
Signature · 1 line
2.0.3
#18 lore raw

RE: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

Quoted 123 lines

From: [email protected] [mailto:[email protected]]
Sent: Tuesday, October 28, 2014 4:26 PM

Since the dwc2 hcd driver is currently not looking for a clock node during init, we should not completely fail if there isn't a clock provided. For dual-role mode, we will only fail init for a non-clock node error. We then update the HCD to only call gadget funtions if there is a proper clock node.

---
v5: reworked to not access gadget functions from the hcd.

---
 drivers/usb/dwc2/core.h      |  3 +--
 drivers/usb/dwc2/core_intr.c |  9 ++++++---
 drivers/usb/dwc2/hcd.c       |  3 ++-
 drivers/usb/dwc2/platform.c  | 19 +++++++++++++++----
 4 files changed, 24 insertions(+), 10 deletions(-)
drivers/usb/dwc2/core.h +1 -2
index ec70862..48120c8 100644
@@ -660,6 +660,7 @@ struct dwc2_hsotg {
 #endif
 #endif /* CONFIG_USB_DWC2_HOST || CONFIG_USB_DWC2_DUAL_ROLE */
 
+	struct clk *clk;
 #if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 	/* Gadget structures */
 	struct usb_gadget_driver *driver;
@@ -667,8 +668,6 @@ struct dwc2_hsotg {
 	struct usb_phy *uphy;
 	struct s3c_hsotg_plat *plat;
 
-	struct clk *clk;
-
 	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
 
 	u32 phyif;
drivers/usb/dwc2/core_intr.c +6 -3
index 1240875..1608037 100644
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);
 	} else {
 		if (hsotg->lx_state != DWC2_L1) {
 			u32 pcgcctl = readl(hsotg->regs + PCGCTL);
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);
 
 	gintsts = dwc2_read_common_intr(hsotg);
 	if (gintsts & ~GINTSTS_PRTINT)
drivers/usb/dwc2/hcd.c +2 -1
index 44c609f..fa49c72 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct work_struct *work)
 		hsotg->op_state = OTG_STATE_B_PERIPHERAL;
 		dwc2_core_init(hsotg, false, -1);
 		dwc2_enable_global_interrupts(hsotg);
-		s3c_hsotg_core_init(hsotg);
+		if (!IS_ERR(hsotg->clk))
+			s3c_hsotg_core_init(hsotg);
 	} else {
 		/* A-Device connector (Host Mode) */
 		dev_dbg(hsotg->dev, "connId A\n");
drivers/usb/dwc2/platform.c +13 -3
index 72f32f7..77c8417 100644
@@ -217,8 +217,17 @@ static int dwc2_driver_probe(struct platform_device *dev)
 
 	spin_lock_init(&hsotg->lock);
 	retval = dwc2_gadget_init(hsotg, irq);
-	if (retval)
-		return retval;
+	if (retval) {
+		/*
+		 * We will not fail the driver initialization for dual-role
+		 * if no clock node is supplied. However, all gadget
+		 * functionality will be disabled if a clock node is not
+		 * provided. Host functionality will continue.
+		 * TO-DO: make clock node a requirement for the HCD.
+		 */
+		if (!IS_ERR(hsotg->clk))
+			return retval;
+	}
 	retval = dwc2_hcd_init(hsotg, irq, params);
 	if (retval)
 		return retval;
@@ -234,7 +243,8 @@ static int dwc2_suspend(struct device *dev)
 	int ret = 0;
 
 	if (dwc2_is_device_mode(dwc2))
-		ret = s3c_hsotg_suspend(dwc2);
+		if (!IS_ERR(dwc2->clk))
+			ret = s3c_hsotg_suspend(dwc2);
 	return ret;
 }
 

drivers/usb/dwc2/platform.c static int dwc2_resume(struct device *dev) · line 261

Quoted 9 lines
drivers/usb/dwc2/platform.c +2 -1
@@ -244,7 +254,8 @@ static int dwc2_resume(struct device *dev)
 	int ret = 0;
 
 	if (dwc2_is_device_mode(dwc2))
-		ret = s3c_hsotg_resume(dwc2);
+		if (!IS_ERR(dwc2->clk))
+			ret = s3c_hsotg_resume(dwc2);
 
 	return ret;
 }

Hi Felipe,

I don't feel qualified to ack/nak this patch, since I don't fully understand the clock subsystem. Can you review this one and see if it looks OK?

Signature · 1 line
Paul
#19 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

In reply to [email protected]

Hi,

Quoted 44 lines · On Tue, Oct 28, 2014 at 06:25:47PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Since the dwc2 hcd driver is currently not looking for a clock node during init, we should not completely fail if there isn't a clock provided. For dual-role mode, we will only fail init for a non-clock node error. We then update the HCD to only call gadget funtions if there is a proper clock node.

---
v5: reworked to not access gadget functions from the hcd.

---
 drivers/usb/dwc2/core.h      |  3 +--
 drivers/usb/dwc2/core_intr.c |  9 ++++++---
 drivers/usb/dwc2/hcd.c       |  3 ++-
 drivers/usb/dwc2/platform.c  | 19 +++++++++++++++----
 4 files changed, 24 insertions(+), 10 deletions(-)
drivers/usb/dwc2/core.h +1 -2
index ec70862..48120c8 100644
@@ -660,6 +660,7 @@ struct dwc2_hsotg {
 #endif
 #endif /* CONFIG_USB_DWC2_HOST || CONFIG_USB_DWC2_DUAL_ROLE */
 
+	struct clk *clk;
 #if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 	/* Gadget structures */
 	struct usb_gadget_driver *driver;
@@ -667,8 +668,6 @@ struct dwc2_hsotg {
 	struct usb_phy *uphy;
 	struct s3c_hsotg_plat *plat;
 
-	struct clk *clk;
-
 	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
 
 	u32 phyif;
drivers/usb/dwc2/core_intr.c +0 -0
index 1240875..1608037 100644

drivers/usb/dwc2/core_intr.c static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg) · line 343

Quoted 6 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);

instead of exposing the clock detail to the entire driver, add IS_ERR() checks to resume and suspend instead. In fact, NULL is a valid clock, so you might as well:

	clk = clk_get(foo, bar);
	if (IS_ERR(clk))
		dwc->clk = NULL;
	else
		dwc->clk = clk;

Then you don't need any IS_ERR() checks sprinkled around the driver.

Quoted 11 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");

drivers/usb/dwc2/core_intr.c irqreturn_t dwc2_handle_common_intr(int irq, void *dev) · line 483

Quoted 6 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);

wait a minute, if there is no clock we don't call the gadget interrupt handler ? Why ? Who will disable the IRQ line ?

Quoted 25 lines
drivers/usb/dwc2/hcd.c +2 -1
index 44c609f..fa49c72 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct work_struct *work)
 		hsotg->op_state = OTG_STATE_B_PERIPHERAL;
 		dwc2_core_init(hsotg, false, -1);
 		dwc2_enable_global_interrupts(hsotg);
-		s3c_hsotg_core_init(hsotg);
+		if (!IS_ERR(hsotg->clk))
+			s3c_hsotg_core_init(hsotg);
 	} else {
 		/* A-Device connector (Host Mode) */
 		dev_dbg(hsotg->dev, "connId A\n");
drivers/usb/dwc2/platform.c +1 -2
index 72f32f7..77c8417 100644
@@ -217,8 +217,17 @@ static int dwc2_driver_probe(struct platform_device *dev)
 
 	spin_lock_init(&hsotg->lock);
 	retval = dwc2_gadget_init(hsotg, irq);
-	if (retval)
-		return retval;
+	if (retval) {

drivers/usb/dwc2/platform.c static int dwc2_driver_probe(struct platform_device *dev) · line 230

Quoted 10 lines
drivers/usb/dwc2/platform.c +10 -0
+		/*
+		 * We will not fail the driver initialization for dual-role
+		 * if no clock node is supplied. However, all gadget
+		 * functionality will be disabled if a clock node is not
+		 * provided. Host functionality will continue.
+		 * TO-DO: make clock node a requirement for the HCD.
+		 */
+		if (!IS_ERR(hsotg->clk))
+			return retval;
+	}

no here... this should have been taken care by dwc2_gadget_init() itself.

Signature · 1 line
balbi
#20 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

Quoted 58 lines · Felipe Balbi

Hi,

Quoted 44 lines · On Tue, Oct 28, 2014 at 06:25:47PM -0500, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Since the dwc2 hcd driver is currently not looking for a clock node during init, we should not completely fail if there isn't a clock provided. For dual-role mode, we will only fail init for a non-clock node error. We then update the HCD to only call gadget funtions if there is a proper clock node.

---
v5: reworked to not access gadget functions from the hcd.

---
 drivers/usb/dwc2/core.h      |  3 +--
 drivers/usb/dwc2/core_intr.c |  9 ++++++---
 drivers/usb/dwc2/hcd.c       |  3 ++-
 drivers/usb/dwc2/platform.c  | 19 +++++++++++++++----
 4 files changed, 24 insertions(+), 10 deletions(-)
drivers/usb/dwc2/core.h +1 -2
index ec70862..48120c8 100644
@@ -660,6 +660,7 @@ struct dwc2_hsotg {
 #endif
 #endif /* CONFIG_USB_DWC2_HOST || CONFIG_USB_DWC2_DUAL_ROLE */
 
+	struct clk *clk;
 #if IS_ENABLED(CONFIG_USB_DWC2_PERIPHERAL) || IS_ENABLED(CONFIG_USB_DWC2_DUAL_ROLE)
 	/* Gadget structures */
 	struct usb_gadget_driver *driver;
@@ -667,8 +668,6 @@ struct dwc2_hsotg {
 	struct usb_phy *uphy;
 	struct s3c_hsotg_plat *plat;
 
-	struct clk *clk;
-
 	struct regulator_bulk_data supplies[ARRAY_SIZE(s3c_hsotg_supply_names)];
 
 	u32 phyif;
drivers/usb/dwc2/core_intr.c +0 -0
index 1240875..1608037 100644

drivers/usb/dwc2/core_intr.c static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg) · line 343

Quoted 6 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);

instead of exposing the clock detail to the entire driver, add IS_ERR() checks to resume and suspend instead. In fact, NULL is a valid clock, so you might as well:

Quoted 7 lines
	clk = clk_get(foo, bar);
	if (IS_ERR(clk))
		dwc->clk = NULL;
	else
		dwc->clk = clk;

Then you don't need any IS_ERR() checks sprinkled around the driver.

But we would still need to check for the clock before accessing gadget functionality right?

	if (dwc2->clk)
		call_gadget();
Quoted 14 lines
Quoted 14 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
Quoted 7 lines

drivers/usb/dwc2/core_intr.c irqreturn_t dwc2_handle_common_intr(int irq, void *dev) · line 483

Quoted 4 lines
drivers/usb/dwc2/core_intr.c +2 -1
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);

wait a minute, if there is no clock we don't call the gadget interrupt handler ? Why ? Who will disable the IRQ line ?

This portion is no longer needed when I use shared IRQ lines.

Quoted 31 lines
Quoted 31 lines
drivers/usb/dwc2/hcd.c +2 -1
index 44c609f..fa49c72 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct work_struct *work)
 		hsotg->op_state = OTG_STATE_B_PERIPHERAL;
 		dwc2_core_init(hsotg, false, -1);
 		dwc2_enable_global_interrupts(hsotg);
-		s3c_hsotg_core_init(hsotg);
+		if (!IS_ERR(hsotg->clk))
+			s3c_hsotg_core_init(hsotg);
 	} else {
 		/* A-Device connector (Host Mode) */
 		dev_dbg(hsotg->dev, "connId A\n");
drivers/usb/dwc2/platform.c +6 -2
index 72f32f7..77c8417 100644
@@ -217,8 +217,17 @@ static int dwc2_driver_probe(struct platform_device *dev)
 
 	spin_lock_init(&hsotg->lock);
 	retval = dwc2_gadget_init(hsotg, irq);
-	if (retval)
-		return retval;
+	if (retval) {
+		/*
+		 * We will not fail the driver initialization for dual-role
+		 * if no clock node is supplied. However, all gadget
+		 * functionality will be disabled if a clock node is not
+		 * provided. Host functionality will continue.
Quoted 8 lines

drivers/usb/dwc2/platform.c static int dwc2_driver_probe(struct platform_device *dev) · line 230

Quoted 5 lines
drivers/usb/dwc2/platform.c +5 -0
+		 * TO-DO: make clock node a requirement for the HCD.
+		 */
+		if (!IS_ERR(hsotg->clk))
+			return retval;
+	}

no here... this should have been taken care by dwc2_gadget_init() itself.

Fixed..

Thanks,
Dinh

#21 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

Hi,

Quoted 18 lines · Dinh Nguyen
Quoted 18 lines

drivers/usb/dwc2/core_intr.c static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg) · line 343

Quoted 7 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);

instead of exposing the clock detail to the entire driver, add IS_ERR() checks to resume and suspend instead. In fact, NULL is a valid clock, so you might as well:

	clk = clk_get(foo, bar);
	if (IS_ERR(clk))
		dwc->clk = NULL;
	else
		dwc->clk = clk;
Quoted 7 lines
Quoted 1 line

Then you don't need any IS_ERR() checks sprinkled around the driver.

But we would still need to check for the clock before accessing gadget functionality right?

	if (dwc2->clk)
		call_gadget();

Read my comment again. "NULL is a valid clock". Look at what clk_enable() does when a NULL pointer is passed:

static int __clk_enable(struct clk *clk)

{
	int ret = 0;

	if (!clk)
		return 0;

	if (WARN_ON(clk->prepare_count == 0))
		return -ESHUTDOWN;

	if (clk->enable_count == 0) {
		ret = __clk_enable(clk->parent);

		if (ret)
			return ret;

		if (clk->ops->enable) {
			ret = clk->ops->enable(clk->hw);
			if (ret) {
				__clk_disable(clk->parent);
				return ret;
			}
		}
	}

	clk->enable_count++;
	return 0;
}

int clk_enable(struct clk *clk)

{
	unsigned long flags;
	int ret;

	flags = clk_enable_lock();
	ret = __clk_enable(clk);
	clk_enable_unlock(flags);

	return ret;
}

EXPORT_SYMBOL_GPL(clk_enable);

Quoted 21 lines
Quoted 13 lines
Quoted 13 lines
drivers/usb/dwc2/core_intr.c +2 -1
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
Quoted 7 lines

drivers/usb/dwc2/core_intr.c irqreturn_t dwc2_handle_common_intr(int irq, void *dev) · line 483

Quoted 4 lines
drivers/usb/dwc2/core_intr.c +2 -1
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);

wait a minute, if there is no clock we don't call the gadget interrupt handler ? Why ? Who will disable the IRQ line ?

Quoted 1 line

This portion is no static int __clk_enable(struct clk *clk)

huh ? What I mean is that this has the potential of leaving that IRQ line enabled. Imagine you don't have a clock and s3c_hsotg_irq() isn't called, then a peripheral IRQ fires, since the handler isn't called, who will clear the interrupt ?

Signature · 1 line
balbi
#22 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

Quoted 72 lines · Felipe Balbi

Hi,

Quoted 18 lines · Dinh Nguyen
Quoted 18 lines

Review static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg) · line 343

Quoted 7 lines
(quoted diff) +2 -1
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);

instead of exposing the clock detail to the entire driver, add IS_ERR() checks to resume and suspend instead. In fact, NULL is a valid clock, so you might as well:

	clk = clk_get(foo, bar);
	if (IS_ERR(clk))
		dwc->clk = NULL;
	else
		dwc->clk = clk;
Quoted 7 lines
Quoted 1 line

Then you don't need any IS_ERR() checks sprinkled around the driver.

But we would still need to check for the clock before accessing gadget functionality right?

	if (dwc2->clk)
		call_gadget();

Read my comment again. "NULL is a valid clock". Look at what clk_enable() does when a NULL pointer is passed:

static int __clk_enable(struct clk *clk)

{
	int ret = 0;

	if (!clk)
		return 0;

	if (WARN_ON(clk->prepare_count == 0))
		return -ESHUTDOWN;

	if (clk->enable_count == 0) {
		ret = __clk_enable(clk->parent);

		if (ret)
			return ret;

		if (clk->ops->enable) {
			ret = clk->ops->enable(clk->hw);
			if (ret) {
				__clk_disable(clk->parent);
				return ret;
			}
		}
	}

	clk->enable_count++;
	return 0;
}

int clk_enable(struct clk *clk)

{
	unsigned long flags;
	int ret;

	flags = clk_enable_lock();
	ret = __clk_enable(clk);
	clk_enable_unlock(flags);

	return ret;
}
Quoted 1 line

EXPORT_SYMBOL_GPL(clk_enable);

Ah yes, thanks for the explanation. So if clk=NULL, it just return 0. But what I'm saying is that if the driver is configured for dual-role mode, and no clock is specified, then the driver should not be accessing any gadget functionality.

So as the patch series stands right now, if I swap out an A connector to a B-connector, then I get a connect_id_status change interrupt. The status would show a device and I would initialize the gadget portion of the driver.

drivers/usb/dwc2/hcd.c +0 -0
index 44c609f..96810f7 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct

work_struct *work)

                hsotg->op_state = OTG_STATE_B_PERIPHERAL;
                dwc2_core_init(hsotg, false, -1);
                dwc2_enable_global_interrupts(hsotg);

- s3c_hsotg_core_init(hsotg);
+ if (hsotg->clk)
+ s3c_hsotg_core_init(hsotg);

So if I don't have a valid clock, I'll be accessing the peripheral portion of the IP.

But I guess not having the check for the valid clock here should be fine as I don't see a case where there can be 2 different clocks for host and peripheral?

Quoted 22 lines
Quoted 22 lines
Quoted 21 lines
Quoted 12 lines
(quoted diff) +2 -1
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");

Review irqreturn_t dwc2_handle_common_intr(int irq, void *dev) · line 483

Quoted 6 lines
(quoted diff) +2 -1
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);

wait a minute, if there is no clock we don't call the gadget interrupt handler ? Why ? Who will disable the IRQ line ?

Quoted 7 lines
Quoted 1 line

This portion is no static int __clk_enable(struct clk *clk)

huh ? What I mean is that this has the potential of leaving that IRQ line enabled. Imagine you don't have a clock and s3c_hsotg_irq() isn't called, then a peripheral IRQ fires, since the handler isn't called, who will clear the interrupt ?

Yes, right. This portion of the code is no longer needed when I switched to use a shared IRQ.

Dinh

#23 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

Quoted 101 lines · Dinh Nguyen
Quoted 72 lines · Felipe Balbi

Hi,

Quoted 18 lines · Dinh Nguyen
Quoted 18 lines

Review static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg) · line 343

Quoted 7 lines
(quoted diff) +2 -1
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);

instead of exposing the clock detail to the entire driver, add IS_ERR() checks to resume and suspend instead. In fact, NULL is a valid clock, so you might as well:

	clk = clk_get(foo, bar);
	if (IS_ERR(clk))
		dwc->clk = NULL;
	else
		dwc->clk = clk;
Quoted 7 lines
Quoted 1 line

Then you don't need any IS_ERR() checks sprinkled around the driver.

But we would still need to check for the clock before accessing gadget functionality right?

	if (dwc2->clk)
		call_gadget();

Read my comment again. "NULL is a valid clock". Look at what clk_enable() does when a NULL pointer is passed:

static int __clk_enable(struct clk *clk)

{
	int ret = 0;

	if (!clk)
		return 0;

	if (WARN_ON(clk->prepare_count == 0))
		return -ESHUTDOWN;

	if (clk->enable_count == 0) {
		ret = __clk_enable(clk->parent);

		if (ret)
			return ret;

		if (clk->ops->enable) {
			ret = clk->ops->enable(clk->hw);
			if (ret) {
				__clk_disable(clk->parent);
				return ret;
			}
		}
	}

	clk->enable_count++;
	return 0;
}

int clk_enable(struct clk *clk)

{
	unsigned long flags;
	int ret;

	flags = clk_enable_lock();
	ret = __clk_enable(clk);
	clk_enable_unlock(flags);

	return ret;
}
Quoted 1 line

EXPORT_SYMBOL_GPL(clk_enable);

Ah yes, thanks for the explanation. So if clk=NULL, it just return 0. But what I'm saying is that if the driver is configured for dual-role mode, and no clock is specified, then the driver should not be accessing any gadget functionality.

So as the patch series stands right now, if I swap out an A connector to a B-connector, then I get a connect_id_status change interrupt. The status would show a device and I would initialize the gadget portion of the driver.

drivers/usb/dwc2/hcd.c +2 -1
index 44c609f..96810f7 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct
 work_struct *work)
                hsotg->op_state = OTG_STATE_B_PERIPHERAL;
                dwc2_core_init(hsotg, false, -1);
                dwc2_enable_global_interrupts(hsotg);
-               s3c_hsotg_core_init(hsotg);
+               if (hsotg->clk)
+                       s3c_hsotg_core_init(hsotg);
 

So if I don't have a valid clock, I'll be accessing the peripheral portion of the IP.

Quoted 4 lines

But I guess not having the check for the valid clock here should be fine as I don't see a case where there can be 2 different clocks for host and peripheral?

Ah...nevermind. I don't need to check for clocks at all because in dwc2_gadget_init(), the clock node check comes before usb_add_gadget_udc(). Thus without a clock node, gadget functionality is disabled already.

Thanks,
Dinh

#24 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

In reply to Dinh Nguyen

Hi,

Quoted 71 lines · Dinh Nguyen
Quoted 70 lines · Felipe Balbi

Hi,

Quoted 18 lines · Dinh Nguyen
Quoted 18 lines

Review static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg) · line 343

Quoted 7 lines
(quoted diff) +2 -1
@@ -339,7 +339,8 @@ static void dwc2_handle_wakeup_detected_intr(struct dwc2_hsotg *hsotg)
 		}
 		/* Change to L0 state */
 		hsotg->lx_state = DWC2_L0;
-		call_gadget(hsotg, resume);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, resume);

instead of exposing the clock detail to the entire driver, add IS_ERR() checks to resume and suspend instead. In fact, NULL is a valid clock, so you might as well:

	clk = clk_get(foo, bar);
	if (IS_ERR(clk))
		dwc->clk = NULL;
	else
		dwc->clk = clk;
Quoted 7 lines
Quoted 1 line

Then you don't need any IS_ERR() checks sprinkled around the driver.

But we would still need to check for the clock before accessing gadget functionality right?

	if (dwc2->clk)
		call_gadget();

Read my comment again. "NULL is a valid clock". Look at what clk_enable() does when a NULL pointer is passed:

static int __clk_enable(struct clk *clk)

{
	int ret = 0;

	if (!clk)
		return 0;

	if (WARN_ON(clk->prepare_count == 0))
		return -ESHUTDOWN;

	if (clk->enable_count == 0) {
		ret = __clk_enable(clk->parent);

		if (ret)
			return ret;

		if (clk->ops->enable) {
			ret = clk->ops->enable(clk->hw);
			if (ret) {
				__clk_disable(clk->parent);
				return ret;
			}
		}
	}

	clk->enable_count++;
	return 0;
}

int clk_enable(struct clk *clk)

{
	unsigned long flags;
	int ret;

	flags = clk_enable_lock();
	ret = __clk_enable(clk);
	clk_enable_unlock(flags);
Quoted 8 lines
Quoted 3 lines
	return ret;
}

EXPORT_SYMBOL_GPL(clk_enable);

Ah yes, thanks for the explanation. So if clk=NULL, it just return 0. But what I'm saying is that if the driver is configured for dual-role mode, and no clock is specified, then the driver should not be accessing any gadget functionality.

why ? Why only for gadget and why can't it work on platforms without clk? What if Paul wants to run the gadget side on his HAPS-5x platform configured as a PCIe card ?

You haven't explained why gadget has this hard-dependency on clk and why *only* gadget has it. This sounds really, really wrong. Why can host side run without clk but gadget can't ?

Moreover, if you really want to prevent people from using gadget without clock, fail dwc2_gadget_init() and have the core fallback to host-only, then print a warning message.

Quoted 4 lines

So as the patch series stands right now, if I swap out an A connector to a B-connector, then I get a connect_id_status change interrupt. The status would show a device and I would initialize the gadget portion of the driver.

that's fine, but why the hard-dependency on clk ?

Quoted 13 lines
drivers/usb/dwc2/hcd.c +2 -1
index 44c609f..96810f7 100644
@@ -1371,7 +1371,8 @@ static void dwc2_conn_id_status_change(struct
 work_struct *work)
                hsotg->op_state = OTG_STATE_B_PERIPHERAL;
                dwc2_core_init(hsotg, false, -1);
                dwc2_enable_global_interrupts(hsotg);
-               s3c_hsotg_core_init(hsotg);
+               if (hsotg->clk)
+                       s3c_hsotg_core_init(hsotg);
 
Quoted 2 lines

So if I don't have a valid clock, I'll be accessing the peripheral portion of the IP.

so what ?

Quoted 3 lines

But I guess not having the check for the valid clock here should be fine as I don't see a case where there can be 2 different clocks for host and peripheral?

probably the same thing.

Quoted 23 lines
Quoted 23 lines
Quoted 22 lines
Quoted 13 lines
Quoted 13 lines
(quoted diff) +2 -1
@@ -400,7 +401,8 @@ static void dwc2_handle_usb_suspend_intr(struct dwc2_hsotg *hsotg)
 			"DSTS.Suspend Status=%d HWCFG4.Power Optimize=%d\n",
 			!!(dsts & DSTS_SUSPSTS),
 			hsotg->hw_params.power_optimized);
-		call_gadget(hsotg, suspend);
+		if (!IS_ERR(hsotg->clk))
+			call_gadget(hsotg, suspend);
 	} else {
 		if (hsotg->op_state == OTG_STATE_A_PERIPHERAL) {
 			dev_dbg(hsotg->dev, "a_peripheral->a_host\n");
@@ -477,7 +479,8 @@ irqreturn_t dwc2_handle_common_intr(int irq, void *dev)
 	spin_lock(&hsotg->lock);
 
Quoted 7 lines

Review irqreturn_t dwc2_handle_common_intr(int irq, void *dev) · line 483

Quoted 4 lines
(quoted diff) +2 -1
 	if (dwc2_is_device_mode(hsotg))
-		retval = s3c_hsotg_irq(irq, dev);
+		if (!IS_ERR(hsotg->clk))
+			retval = s3c_hsotg_irq(irq, dev);

wait a minute, if there is no clock we don't call the gadget interrupt handler ? Why ? Who will disable the IRQ line ?

This portion is no static int __clk_enable(struct clk *clk)

Quoted 8 lines
Quoted 5 lines

huh ? What I mean is that this has the potential of leaving that IRQ line enabled. Imagine you don't have a clock and s3c_hsotg_irq() isn't called, then a peripheral IRQ fires, since the handler isn't called, who will clear the interrupt ?

Yes, right. This portion of the code is no longer needed when I switched to use a shared IRQ.

k

Signature · 1 line
balbi
#25 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

In reply to [email protected]

Hi Dinh,

Quoted 7 lines · On 10/29/2014 07:25 AM, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Since the dwc2 hcd driver is currently not looking for a clock node during init, we should not completely fail if there isn't a clock provided. For dual-role mode, we will only fail init for a non-clock node error. We then update the HCD to only call gadget funtions if there is a proper clock node.

We have to add clock management for hcd, and I think it is better to do it before more Socs use this driver, isn't it?
I have do something in my RFC patches, but I think I still do it in a wrong way. Can we just handle all the clock thing in platform?

Balbi suggested in my patch that we can "hide" clk_enable()/disable() under ->runtime_resume()/->runtime_suspend() and linux driver model. Can this be in platform driver?

- Kever

#26 lore raw

Re: [PATCHv6 6/8] usb: dwc2: gadget: Do not fail probe if there isn't a clock node

Quoted 16 lines · Kever Yang

Hi Dinh,

Quoted 7 lines · On 10/29/2014 07:25 AM, [email protected] wrote:

From: Dinh Nguyen <[email protected]>

Since the dwc2 hcd driver is currently not looking for a clock node during init, we should not completely fail if there isn't a clock provided. For dual-role mode, we will only fail init for a non-clock node error. We then update the HCD to only call gadget funtions if there is a proper clock node.

We have to add clock management for hcd, and I think it is better to do it before more Socs use this driver, isn't it?
I have do something in my RFC patches, but I think I still do it in a wrong way.
Can we just handle all the clock thing in platform?

Quoted 3 lines

Balbi suggested in my patch that we can "hide" clk_enable()/disable() under ->runtime_resume()/->runtime_suspend() and linux driver model. Can this be in platform driver?

it can and it probably should. Implement ->runtime_resume()/->runtime_suspend()/runtime_idle() in platform.c and call pm_runtime_enable()/get()/put()/mark_last_busy()/autosuspend() properly.

Signature · 1 line
balbi
#27 lore raw

PATCH v6 7/8 usb: dwc2: Update Kconfig to support dual-role

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

Update DWC2 kconfig and makefile to support dual-role mode. The platform file will always get compiled for the case where the controller is directly connected to the CPU. So for loadable modules, dwc2.ko is built for host, peripheral, and dual-role mode. The PCI bus interface will be called dwc2_pci.ko and the platform interface module will be called dwc2_platform.ko.

---
v6: Correct notes for USB_DWC2_DUAL_ROLE and USB_DWC2_PERIPHERAL
Remove extra default for USB_DWC2_PLATFORM and make it a condition for building dwc2_platform.ko. In addition USB_DWC2_PLATFORM and USB_DWC2_PCI are now tristate.
v5: dwc2.ko for all modes along with dwc2_platform.ko and dwc2_pci.ko will
get built for kernel modules.
v3: Add USB_GADGET=y and USB_GADGET=USB_DWC2 for peripheral and dual-role
config options.
v2: Remove reference to dwc2_gadget

---
 drivers/usb/dwc2/Kconfig  | 66 ++++++++++++++++++++++++++++-------------------
 drivers/usb/dwc2/Makefile | 32 +++++++++++------------
 2 files changed, 55 insertions(+), 43 deletions(-)
drivers/usb/dwc2/Kconfig +39 -27
index 4d02718..b323c4c 100644
@@ -1,5 +1,5 @@
 config USB_DWC2
-	bool "DesignWare USB2 DRD Core Support"
+	tristate "DesignWare USB2 DRD Core Support"
 	depends on USB || USB_GADGET
 	help
 	  Say Y here if your system has a Dual Role Hi-Speed USB
@@ -10,49 +10,61 @@ config USB_DWC2
 	  bus interface module (if you have a PCI bus system) will be
 	  called dwc2_pci.ko, and the platform interface module (for
 	  controllers directly connected to the CPU) will be called
-	  dwc2_platform.ko. For gadget mode, there will be a single
-	  module called dwc2_gadget.ko.
-
-	  NOTE: The s3c-hsotg driver is now renamed to dwc2_gadget. The
-	  host and gadget drivers are still currently separate drivers.
-	  There are plans to merge the dwc2_gadget driver with the dwc2
-	  host driver in the near future to create a dual-role driver.
+	  dwc2_platform.ko. For all modes(host, gadget and dual-role), there
+	  will be an additional module named dwc2.ko.
 
 if USB_DWC2
 
+choice
+	bool "DWC2 Mode Selection"
+	default USB_DWC2_DUAL_ROLE if (USB && USB_GADGET)
+	default USB_DWC2_HOST if (USB && !USB_GADGET)
+	default USB_DWC2_PERIPHERAL if (!USB && USB_GADGET)
+
 config USB_DWC2_HOST
-	tristate "Host only mode"
+	bool "Host only mode"
 	depends on USB
 	help
 	  The Designware USB2.0 high-speed host controller
-	  integrated into many SoCs.
+	  integrated into many SoCs. Select this option if you want the
+	  driver to operate in Host-only mode.
 
-config USB_DWC2_PLATFORM
-	bool "DWC2 Platform"
-	depends on USB_DWC2_HOST
-	default USB_DWC2_HOST
+comment "Gadget/Dual-role mode requires USB Gadget support to be enabled"
+
+config USB_DWC2_PERIPHERAL
+	bool "Gadget only mode"
+	depends on USB_GADGET=y || USB_GADGET=USB_DWC2
+	help
+	  The Designware USB2.0 high-speed gadget controller
+	  integrated into many SoCs. Select this option if you want the
+	  driver to operate in Peripheral-only mode. This option requires
+	  USB_GADGET to be enabled.
+
+config USB_DWC2_DUAL_ROLE
+	bool "Dual Role mode"
+	depends on (USB=y || USB=USB_DWC2) && (USB_GADGET=y || USB_GADGET=USB_DWC2)
 	help
-	  The Designware USB2.0 platform interface module for
-	  controllers directly connected to the CPU. This is only
-	  used for host mode.
+	  Select this option if you want the driver to work in a dual-role
+	  mode. In this mode both host and gadget features are enabled, and
+	  the role will be determined by the cable that gets plugged-in. This
+	  option requires USB_GADGET to be enabled.
+endchoice
+
+config USB_DWC2_PLATFORM
+	tristate "DWC2 Platform"
+	default USB_DWC2_HOST || USB_DWC2_PERIPHERAL
+        help
+          The Designware USB2.0 platform interface module for
+          controllers directly connected to the CPU.
 
 config USB_DWC2_PCI
-	bool "DWC2 PCI"
+	tristate "DWC2 PCI"
 	depends on USB_DWC2_HOST && PCI
 	default USB_DWC2_HOST
 	help
 	  The Designware USB2.0 PCI interface module for controllers
 	  connected to a PCI bus. This is only used for host mode.
 
-comment "Gadget mode requires USB Gadget support to be enabled"
-
-config USB_DWC2_PERIPHERAL
-	tristate "Gadget only mode"
-	depends on USB_GADGET
-	help
-	  The Designware USB2.0 high-speed gadget controller
-	  integrated into many SoCs.
-
 config USB_DWC2_DEBUG
 	bool "Enable Debugging Messages"
 	help
drivers/usb/dwc2/Makefile +16 -16
index b73d2a5..8f75267 100644
@@ -1,28 +1,28 @@
 ccflags-$(CONFIG_USB_DWC2_DEBUG)	+= -DDEBUG
 ccflags-$(CONFIG_USB_DWC2_VERBOSE)	+= -DVERBOSE_DEBUG
 
-obj-$(CONFIG_USB_DWC2_HOST)		+= dwc2.o
+obj-$(CONFIG_USB_DWC2)			+= dwc2.o
 dwc2-y					:= core.o core_intr.o
-dwc2-y					+= hcd.o hcd_intr.o
-dwc2-y					+= hcd_queue.o hcd_ddma.o
+
+ifneq ($(filter y,$(CONFIG_USB_DWC2_HOST) $(CONFIG_USB_DWC2_DUAL_ROLE)),)
+	dwc2-y				+= hcd.o hcd_intr.o
+	dwc2-y				+= hcd_queue.o hcd_ddma.o
+endif
+
+ifneq ($(filter y,$(CONFIG_USB_DWC2_PERIPHERAL) $(CONFIG_USB_DWC2_DUAL_ROLE)),)
+	dwc2-y       			+= gadget.o
+endif
 
 # NOTE: The previous s3c-hsotg peripheral mode only driver has been moved to
 # this location and renamed gadget.c. When building for dynamically linked
-# modules, dwc2_gadget.ko will get built for peripheral mode. For host mode,
-# the core module will be dwc2.ko, the PCI bus interface module will called
-# dwc2_pci.ko and the platform interface module will be called dwc2_platform.ko.
-# At present the host and gadget driver will be separate drivers, but there
-# are plans in the near future to create a dual-role driver.
+# modules, dwc2.ko will get built for host mode, peripheral mode, and dual-role
+# mode. The PCI bus interface module will called dwc2_pci.ko and the platform
+# interface module will be called dwc2_platform.ko.
 
 ifneq ($(CONFIG_USB_DWC2_PCI),)
-	obj-$(CONFIG_USB_DWC2_HOST)	+= dwc2_pci.o
+	obj-$(CONFIG_USB_DWC2)		+= dwc2_pci.o
 	dwc2_pci-y			:= pci.o
 endif
 
-ifneq ($(CONFIG_USB_DWC2_PLATFORM),)
-	obj-$(CONFIG_USB_DWC2_HOST)	+= dwc2_platform.o
-	dwc2_platform-y			:= platform.o
-endif
-
-obj-$(CONFIG_USB_DWC2_PERIPHERAL)	+= dwc2_gadget.o
-dwc2_gadget-y				:= gadget.o
+obj-$(CONFIG_USB_DWC2_PLATFORM)		+= dwc2_platform.o
+dwc2_platform-y				:= platform.o
Signature · 1 line
2.0.3
#28 lore raw

PATCH v6 8/8 usb: dwc2: move usb_disabled() call to host driver only

In reply to [email protected]

From: Dinh Nguyen <[email protected]>

Now that platform.c will get built for both Host and Gadget, if we leave the usb_disabled() call in platform.c, it results in the following build error when (!USB && USB_GADGET) condition is met.

ERROR: "usb_disabled" [drivers/usb/dwc2/dwc2_platform.ko] undefined!

Since usb_disabled() is mostly used to disable USB host functionality, move the call the host portion for the DWC2 driver.

---
 drivers/usb/dwc2/hcd.c      | 3 +++
 drivers/usb/dwc2/platform.c | 3 ---
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/usb/dwc2/hcd.c +3 -0
index fa49c72..b741997 100644
@@ -2780,6 +2780,9 @@ int dwc2_hcd_init(struct dwc2_hsotg *hsotg, int irq,
 	int i, num_channels;
 	int retval;
 
+	if (usb_disabled())
+		return -ENODEV;
+
 	dev_dbg(hsotg->dev, "DWC OTG HCD INIT\n");
 
 	/* Detect config values from hardware */
drivers/usb/dwc2/platform.c +0 -3
index 77c8417..123cf54 100644
@@ -157,9 +157,6 @@ static int dwc2_driver_probe(struct platform_device *dev)
 	int retval;
 	int irq;
 
-	if (usb_disabled())
-		return -ENODEV;
-
 	match = of_match_device(dwc2_of_match_table, &dev->dev);
 	if (match && match->data) {
 		params = match->data;
Signature · 1 line
2.0.3
#29 lore raw

RE: [PATCHv6 8/8] usb: dwc2: move usb_disabled() call to host driver only

Quoted 37 lines

From: [email protected] [mailto:[email protected]]
Sent: Tuesday, October 28, 2014 4:26 PM

Now that platform.c will get built for both Host and Gadget, if we leave the usb_disabled() call in platform.c, it results in the following build error when (!USB && USB_GADGET) condition is met.

ERROR: "usb_disabled" [drivers/usb/dwc2/dwc2_platform.ko] undefined!

Since usb_disabled() is mostly used to disable USB host functionality, move the call the host portion for the DWC2 driver.

---
 drivers/usb/dwc2/hcd.c      | 3 +++
 drivers/usb/dwc2/platform.c | 3 ---
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/usb/dwc2/hcd.c +3 -0
index fa49c72..b741997 100644
@@ -2780,6 +2780,9 @@ int dwc2_hcd_init(struct dwc2_hsotg *hsotg, int irq,
 	int i, num_channels;
 	int retval;
 
+	if (usb_disabled())
+		return -ENODEV;
+
 	dev_dbg(hsotg->dev, "DWC OTG HCD INIT\n");
 
 	/* Detect config values from hardware */
drivers/usb/dwc2/platform.c +0 -0
index 77c8417..123cf54 100644

drivers/usb/dwc2/platform.c static int dwc2_driver_probe(struct platform_device *dev) · line 162

Quoted 9 lines
drivers/usb/dwc2/platform.c +0 -3
@@ -157,9 +157,6 @@ static int dwc2_driver_probe(struct platform_device *dev)
 	int retval;
 	int irq;
 
-	if (usb_disabled())
-		return -ENODEV;
-
 	match = of_match_device(dwc2_of_match_table, &dev->dev);
 	if (match && match->data) {
 		params = match->data;

I'm confused. You are saying the build is broken until patch 8/8 is applied? As always, that is not acceptable. You need to fix the breakage at the point where it was introduced, not leave it broken until the last patch in the series.

Signature · 1 line
Paul
#30 lore raw

Re: [PATCHv6 8/8] usb: dwc2: move usb_disabled() call to host driver only

Quoted 40 lines · Paul Zimmerman
Quoted 40 lines

From: [email protected] [mailto:[email protected]]
Sent: Tuesday, October 28, 2014 4:26 PM

Now that platform.c will get built for both Host and Gadget, if we leave the usb_disabled() call in platform.c, it results in the following build error when (!USB && USB_GADGET) condition is met.

ERROR: "usb_disabled" [drivers/usb/dwc2/dwc2_platform.ko] undefined!

Since usb_disabled() is mostly used to disable USB host functionality, move the call the host portion for the DWC2 driver.

---
 drivers/usb/dwc2/hcd.c      | 3 +++
 drivers/usb/dwc2/platform.c | 3 ---
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/usb/dwc2/hcd.c +3 -0
index fa49c72..b741997 100644
@@ -2780,6 +2780,9 @@ int dwc2_hcd_init(struct dwc2_hsotg *hsotg, int irq,
 	int i, num_channels;
 	int retval;
 
+	if (usb_disabled())
+		return -ENODEV;
+
 	dev_dbg(hsotg->dev, "DWC OTG HCD INIT\n");
 
 	/* Detect config values from hardware */
drivers/usb/dwc2/platform.c +0 -0
index 77c8417..123cf54 100644
@@ -157,9 +157,6 @@ static int dwc2_driver_probe(struct platform_device *dev)
 	int retval;
 	int irq;
 
Quoted 12 lines

drivers/usb/dwc2/platform.c static int dwc2_driver_probe(struct platform_device *dev) · line 162

Quoted 6 lines
drivers/usb/dwc2/platform.c +0 -3
-	if (usb_disabled())
-		return -ENODEV;
-
 	match = of_match_device(dwc2_of_match_table, &dev->dev);
 	if (match && match->data) {
 		params = match->data;

I'm confused. You are saying the build is broken until patch 8/8 is applied? As always, that is not acceptable. You need to fix the breakage at the point where it was introduced, not leave it broken until the last patch in the series.

The build gets broken when patch 7/8 of is applied. That is the patch that finally allows platform.c to get built for host and gadget. I can fold this patch into patch 7/8.

Dinh

#31 lore raw

Re: [PATCHv6 8/8] usb: dwc2: move usb_disabled() call to host driver only

Quoted 56 lines · Dinh Nguyen
Quoted 40 lines · Paul Zimmerman
Quoted 40 lines

From: [email protected] [mailto:[email protected]]
Sent: Tuesday, October 28, 2014 4:26 PM

Now that platform.c will get built for both Host and Gadget, if we leave the usb_disabled() call in platform.c, it results in the following build error when (!USB && USB_GADGET) condition is met.

ERROR: "usb_disabled" [drivers/usb/dwc2/dwc2_platform.ko] undefined!

Since usb_disabled() is mostly used to disable USB host functionality, move the call the host portion for the DWC2 driver.

---
 drivers/usb/dwc2/hcd.c      | 3 +++
 drivers/usb/dwc2/platform.c | 3 ---
 2 files changed, 3 insertions(+), 3 deletions(-)
drivers/usb/dwc2/hcd.c +3 -0
index fa49c72..b741997 100644
@@ -2780,6 +2780,9 @@ int dwc2_hcd_init(struct dwc2_hsotg *hsotg, int irq,
 	int i, num_channels;
 	int retval;
 
+	if (usb_disabled())
+		return -ENODEV;
+
 	dev_dbg(hsotg->dev, "DWC OTG HCD INIT\n");
 
 	/* Detect config values from hardware */
drivers/usb/dwc2/platform.c +0 -0
index 77c8417..123cf54 100644
@@ -157,9 +157,6 @@ static int dwc2_driver_probe(struct platform_device *dev)
 	int retval;
 	int irq;
 
Quoted 12 lines

drivers/usb/dwc2/platform.c static int dwc2_driver_probe(struct platform_device *dev) · line 162

Quoted 6 lines
drivers/usb/dwc2/platform.c +0 -3
-	if (usb_disabled())
-		return -ENODEV;
-
 	match = of_match_device(dwc2_of_match_table, &dev->dev);
 	if (match && match->data) {
 		params = match->data;

I'm confused. You are saying the build is broken until patch 8/8 is applied? As always, that is not acceptable. You need to fix the breakage at the point where it was introduced, not leave it broken until the last patch in the series.

Quoted 3 lines

The build gets broken when patch 7/8 of is applied. That is the patch that finally allows platform.c to get built for host and gadget. I can fold this patch into patch 7/8.

then you invert things, make $subject patch 7 (or even patch 1) and patch 7 becomes patch 8.

Signature · 1 line
balbi