linux-usb.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [PATCH v12 0/6] Add function suspend/resume and remote wakeup support
@ 2023-03-16 23:38 Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 1/6] usb: gadget: Properly configure the device for remote wakeup Elson Roy Serrao
                   ` (5 more replies)
  0 siblings, 6 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

Changes in v12
 - Added a new patch to arm the function for triggering remote wakeup based on the
   function suspend packet sent by the host. Also handle get_status in composite
   layer as default.
 - Removed set/reset of func_wakeup_armed flag in f_ecm and let composite layer
   handle it
 
Changes in v11
 - When ecm function suspend is called return negative error only if host is trying
   to arm for function wakeup and device is not wakeup capable.
   
Changes in v10
 - Modified the return value to 0 in ecm_get_status() so that device responds
   with both remote wakeup armed and remote wakeup capable bit reset to 0.
 - Return negative errno if wakeup is not supported when func suspend feature
   selector is sent by the host.

Changes in v9
 - Added bmAtrributes wakeup bit check for arming the function for function
   remote wakeup and also in get_status api

Changes in v8
 - Added else case to return error value while setting remote wakeup feature
   selector so that device will respond with a protocl stall

Changes in v7
 - Added a check to set device remote wakeup feature selector in ep0.c based on whether
   the device is configured for remote wakeup.
 - Commit message and usb_func_wakeup documentation changes.

Changes in v6
 - Combined usb_gadget_func_wakeup API with usb_func_wakeup API in composite layer
   so that there is only 1 API for triggering function remote wakeup for better error
   handling. Since function suspend is something specific to usb functions, better to
   keep the related APIs in composite layer and above. Also documented the usage and
   applicability of the usb_func_wakeup API.

Changes in v5
 - Add wakeup_armed check in patch2 in the link status change event handler
   so that resume gets triggeed only in the remote wakeup context.
 - Costmetic changes in patch3 and patch4

Changes in v4
 - Moved the wakeup bit check to bind function for warning the user at an early
   stage itself.
 - Added the remote wakeup configured check to gadget_wakeup() and func_wakeup()
   routines so that wakeup can be triggered only if user has configured it.
 - Cosmetic changes with respect to renaming the variables to reflect the operation
   better.

Changes in v3
 - Modified rw_capable flag to reflect the gadgets capability for wakeup
   signalling.
 - Added a check to configure wakeup bit in bmAttributes only if gadget
   is capable of triggering wakeup.
 - Implemented a gadget op for composite layer to inform UDC whether device
   is configured for remote wakeup.
 - Added a check in __usb_gadget_wakeup() API to trigger wakeup only if the
   device is configured for it.
 - Cosmetic changes in dwc3_gadget_func_wakeup() API.

Changes in v2
 - Added a flag to indicate whether the device is remote wakeup capable.
 - Added an async parameter to _dwc3_gadget_wakeup() API and few cosmetic
   changes.
 - Added flags to reflect the state of  function suspend and function remote
   wakeup to usb_function struct rather than function specific struct (f_ecm).
 - Changed the dwc3_gadget_func__wakeup() API to run synchronously by first
   checking the link state and then sending the device notification. Also
   added debug log for DEVICE_NOTIFICATION generic cmd.
 - Added changes to arm the device for remotewakeup/function remotewakeup
   only if device is capable.

An usb device can initate a remote wakeup and bring the link out of
suspend as dictated by the DEVICE_REMOTE_WAKEUP feature selector.
To achieve this an interface can invoke gadget_wakeup op and wait for the
device to come out of LPM. But the current polling based implementation
fails if the host takes a long time to drive the resume signaling specially
in high speed capable devices. Switching to an interrupt based approach is
more robust and efficient. This can be leveraged by enabling link status
change events and triggering a gadget resume when the link comes to active
state.

If the device is enhanced super-speed capable, individual interfaces can
also be put into suspend state. An interface can be in function suspend
state even when the device is not in suspend state. Function suspend state
is retained throughout the device suspend entry and exit process.
A function can be put to function suspend through FUNCTION_SUSPEND feature
selector sent by the host. This setup packet also decides whether that
function is capable of initiating a function remote wakeup. When the
function sends a wakeup notification to the host the link must be first
brought to a non-U0 state and then this notification is sent.

This change adds the infrastructure needed to support the above
functionalities.

Elson Roy Serrao (6):
  usb: gadget: Properly configure the device for remote wakeup
  usb: dwc3: Add remote wakeup handling
  usb: gadget: Add function wakeup support
  usb: dwc3: Add function suspend and function wakeup support
  usb: gadget: arm the function for triggering remote wakeup
  usb: gadget: f_ecm: Add suspend/resume and remote wakeup support

 drivers/usb/dwc3/core.h               |   5 ++
 drivers/usb/dwc3/debug.h              |   2 +
 drivers/usb/dwc3/ep0.c                |  19 +++---
 drivers/usb/dwc3/gadget.c             | 118 ++++++++++++++++++++++++++++++++--
 drivers/usb/gadget/composite.c        |  84 +++++++++++++++++++++++-
 drivers/usb/gadget/configfs.c         |   3 +
 drivers/usb/gadget/function/f_ecm.c   |  71 ++++++++++++++++++++
 drivers/usb/gadget/function/u_ether.c |  63 ++++++++++++++++++
 drivers/usb/gadget/function/u_ether.h |   4 ++
 drivers/usb/gadget/udc/core.c         |  27 ++++++++
 drivers/usb/gadget/udc/trace.h        |   5 ++
 include/linux/usb/composite.h         |   8 +++
 include/linux/usb/gadget.h            |   9 +++
 13 files changed, 401 insertions(+), 17 deletions(-)

-- 
2.7.4


^ permalink raw reply	[flat|nested] 18+ messages in thread

* [PATCH v12 1/6] usb: gadget: Properly configure the device for remote wakeup
  2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
@ 2023-03-16 23:38 ` Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 2/6] usb: dwc3: Add remote wakeup handling Elson Roy Serrao
                   ` (4 subsequent siblings)
  5 siblings, 0 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

The wakeup bit in the bmAttributes field indicates whether the device
is configured for remote wakeup. But this field should be allowed to
set only if the UDC supports such wakeup mechanism. So configure this
field based on UDC capability. Also inform the UDC whether the device
is configured for remote wakeup by implementing a gadget op.

Reviewed-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
---
 drivers/usb/gadget/composite.c | 18 ++++++++++++++++++
 drivers/usb/gadget/configfs.c  |  3 +++
 drivers/usb/gadget/udc/core.c  | 27 +++++++++++++++++++++++++++
 drivers/usb/gadget/udc/trace.h |  5 +++++
 include/linux/usb/composite.h  |  2 ++
 include/linux/usb/gadget.h     |  8 ++++++++
 6 files changed, 63 insertions(+)

diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
index 36add18..c9c983e 100644
--- a/drivers/usb/gadget/composite.c
+++ b/drivers/usb/gadget/composite.c
@@ -513,6 +513,19 @@ static u8 encode_bMaxPower(enum usb_device_speed speed,
 		return min(val, 900U) / 8;
 }
 
+void check_remote_wakeup_config(struct usb_gadget *g,
+				struct usb_configuration *c)
+{
+	if (USB_CONFIG_ATT_WAKEUP & c->bmAttributes) {
+		/* Reset the rw bit if gadget is not capable of it */
+		if (!g->wakeup_capable && g->ops->set_remote_wakeup) {
+			WARN(c->cdev, "Clearing wakeup bit for config c.%d\n",
+			     c->bConfigurationValue);
+			c->bmAttributes &= ~USB_CONFIG_ATT_WAKEUP;
+		}
+	}
+}
+
 static int config_buf(struct usb_configuration *config,
 		enum usb_device_speed speed, void *buf, u8 type)
 {
@@ -994,6 +1007,11 @@ static int set_config(struct usb_composite_dev *cdev,
 		power = min(power, 500U);
 	else
 		power = min(power, 900U);
+
+	if (USB_CONFIG_ATT_WAKEUP & c->bmAttributes)
+		usb_gadget_set_remote_wakeup(gadget, 1);
+	else
+		usb_gadget_set_remote_wakeup(gadget, 0);
 done:
 	if (power <= USB_SELF_POWER_VBUS_MAX_DRAW)
 		usb_gadget_set_selfpowered(gadget);
diff --git a/drivers/usb/gadget/configfs.c b/drivers/usb/gadget/configfs.c
index b9f1136..4c639e9 100644
--- a/drivers/usb/gadget/configfs.c
+++ b/drivers/usb/gadget/configfs.c
@@ -1761,6 +1761,9 @@ static int configfs_composite_bind(struct usb_gadget *gadget,
 		if (gadget_is_otg(gadget))
 			c->descriptors = otg_desc;
 
+		/* Properly configure the bmAttributes wakeup bit */
+		check_remote_wakeup_config(gadget, c);
+
 		cfg = container_of(c, struct config_usb_cfg, c);
 		if (!list_empty(&cfg->string_list)) {
 			i = 0;
diff --git a/drivers/usb/gadget/udc/core.c b/drivers/usb/gadget/udc/core.c
index 23b0629..3dcbba7 100644
--- a/drivers/usb/gadget/udc/core.c
+++ b/drivers/usb/gadget/udc/core.c
@@ -514,6 +514,33 @@ int usb_gadget_wakeup(struct usb_gadget *gadget)
 EXPORT_SYMBOL_GPL(usb_gadget_wakeup);
 
 /**
+ * usb_gadget_set_remote_wakeup - configures the device remote wakeup feature.
+ * @gadget:the device being configured for remote wakeup
+ * @set:value to be configured.
+ *
+ * set to one to enable remote wakeup feature and zero to disable it.
+ *
+ * returns zero on success, else negative errno.
+ */
+int usb_gadget_set_remote_wakeup(struct usb_gadget *gadget, int set)
+{
+	int ret = 0;
+
+	if (!gadget->ops->set_remote_wakeup) {
+		ret = -EOPNOTSUPP;
+		goto out;
+	}
+
+	ret = gadget->ops->set_remote_wakeup(gadget, set);
+
+out:
+	trace_usb_gadget_set_remote_wakeup(gadget, ret);
+
+	return ret;
+}
+EXPORT_SYMBOL_GPL(usb_gadget_set_remote_wakeup);
+
+/**
  * usb_gadget_set_selfpowered - sets the device selfpowered feature.
  * @gadget:the device being declared as self-powered
  *
diff --git a/drivers/usb/gadget/udc/trace.h b/drivers/usb/gadget/udc/trace.h
index abdbcb1..a5ed26f 100644
--- a/drivers/usb/gadget/udc/trace.h
+++ b/drivers/usb/gadget/udc/trace.h
@@ -91,6 +91,11 @@ DEFINE_EVENT(udc_log_gadget, usb_gadget_wakeup,
 	TP_ARGS(g, ret)
 );
 
+DEFINE_EVENT(udc_log_gadget, usb_gadget_set_remote_wakeup,
+	TP_PROTO(struct usb_gadget *g, int ret),
+	TP_ARGS(g, ret)
+);
+
 DEFINE_EVENT(udc_log_gadget, usb_gadget_set_selfpowered,
 	TP_PROTO(struct usb_gadget *g, int ret),
 	TP_ARGS(g, ret)
diff --git a/include/linux/usb/composite.h b/include/linux/usb/composite.h
index 608dc96..d949e91 100644
--- a/include/linux/usb/composite.h
+++ b/include/linux/usb/composite.h
@@ -413,6 +413,8 @@ extern int composite_dev_prepare(struct usb_composite_driver *composite,
 extern int composite_os_desc_req_prepare(struct usb_composite_dev *cdev,
 					 struct usb_ep *ep0);
 void composite_dev_cleanup(struct usb_composite_dev *cdev);
+void check_remote_wakeup_config(struct usb_gadget *g,
+				struct usb_configuration *c);
 
 static inline struct usb_composite_driver *to_cdriver(
 		struct usb_gadget_driver *gdrv)
diff --git a/include/linux/usb/gadget.h b/include/linux/usb/gadget.h
index 00750f7..1d79612 100644
--- a/include/linux/usb/gadget.h
+++ b/include/linux/usb/gadget.h
@@ -310,6 +310,7 @@ struct usb_udc;
 struct usb_gadget_ops {
 	int	(*get_frame)(struct usb_gadget *);
 	int	(*wakeup)(struct usb_gadget *);
+	int	(*set_remote_wakeup)(struct usb_gadget *, int set);
 	int	(*set_selfpowered) (struct usb_gadget *, int is_selfpowered);
 	int	(*vbus_session) (struct usb_gadget *, int is_active);
 	int	(*vbus_draw) (struct usb_gadget *, unsigned mA);
@@ -384,6 +385,8 @@ struct usb_gadget_ops {
  * @connected: True if gadget is connected.
  * @lpm_capable: If the gadget max_speed is FULL or HIGH, this flag
  *	indicates that it supports LPM as per the LPM ECN & errata.
+ * @wakeup_capable: True if gadget is capable of sending remote wakeup.
+ * @wakeup_armed: True if gadget is armed by the host for remote wakeup.
  * @irq: the interrupt number for device controller.
  * @id_number: a unique ID number for ensuring that gadget names are distinct
  *
@@ -445,6 +448,8 @@ struct usb_gadget {
 	unsigned			deactivated:1;
 	unsigned			connected:1;
 	unsigned			lpm_capable:1;
+	unsigned			wakeup_capable:1;
+	unsigned			wakeup_armed:1;
 	int				irq;
 	int				id_number;
 };
@@ -601,6 +606,7 @@ static inline int gadget_is_otg(struct usb_gadget *g)
 #if IS_ENABLED(CONFIG_USB_GADGET)
 int usb_gadget_frame_number(struct usb_gadget *gadget);
 int usb_gadget_wakeup(struct usb_gadget *gadget);
+int usb_gadget_set_remote_wakeup(struct usb_gadget *gadget, int set);
 int usb_gadget_set_selfpowered(struct usb_gadget *gadget);
 int usb_gadget_clear_selfpowered(struct usb_gadget *gadget);
 int usb_gadget_vbus_connect(struct usb_gadget *gadget);
@@ -616,6 +622,8 @@ static inline int usb_gadget_frame_number(struct usb_gadget *gadget)
 { return 0; }
 static inline int usb_gadget_wakeup(struct usb_gadget *gadget)
 { return 0; }
+static inline int usb_gadget_set_remote_wakeup(struct usb_gadget *gadget, int set)
+{ return 0; }
 static inline int usb_gadget_set_selfpowered(struct usb_gadget *gadget)
 { return 0; }
 static inline int usb_gadget_clear_selfpowered(struct usb_gadget *gadget)
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 18+ messages in thread

* [PATCH v12 2/6] usb: dwc3: Add remote wakeup handling
  2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 1/6] usb: gadget: Properly configure the device for remote wakeup Elson Roy Serrao
@ 2023-03-16 23:38 ` Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 3/6] usb: gadget: Add function wakeup support Elson Roy Serrao
                   ` (3 subsequent siblings)
  5 siblings, 0 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

An usb device can initate a remote wakeup and bring the link out of
suspend as dictated by the DEVICE_REMOTE_WAKEUP feature selector.
Add support to handle this packet and set the remote wakeup capability.

Some hosts may take longer time to initiate the resume signaling after
device triggers a remote wakeup. So add async support to the wakeup API
by enabling link status change events.

Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
---
 drivers/usb/dwc3/core.h   |  2 ++
 drivers/usb/dwc3/ep0.c    |  7 +++++
 drivers/usb/dwc3/gadget.c | 76 +++++++++++++++++++++++++++++++++++++++++++----
 3 files changed, 79 insertions(+), 6 deletions(-)

diff --git a/drivers/usb/dwc3/core.h b/drivers/usb/dwc3/core.h
index b1bd631..416e0ef 100644
--- a/drivers/usb/dwc3/core.h
+++ b/drivers/usb/dwc3/core.h
@@ -1113,6 +1113,7 @@ struct dwc3_scratchpad_array {
  *	3	- Reserved
  * @dis_metastability_quirk: set to disable metastability quirk.
  * @dis_split_quirk: set to disable split boundary.
+ * @wakeup_configured: set if the device is configured for remote wakeup.
  * @imod_interval: set the interrupt moderation interval in 250ns
  *			increments or 0 to disable.
  * @max_cfg_eps: current max number of IN eps used across all USB configs.
@@ -1331,6 +1332,7 @@ struct dwc3 {
 
 	unsigned		dis_split_quirk:1;
 	unsigned		async_callbacks:1;
+	unsigned		wakeup_configured:1;
 
 	u16			imod_interval;
 
diff --git a/drivers/usb/dwc3/ep0.c b/drivers/usb/dwc3/ep0.c
index 61de693..8aa14a5 100644
--- a/drivers/usb/dwc3/ep0.c
+++ b/drivers/usb/dwc3/ep0.c
@@ -356,6 +356,9 @@ static int dwc3_ep0_handle_status(struct dwc3 *dwc,
 				usb_status |= 1 << USB_DEV_STAT_U1_ENABLED;
 			if (reg & DWC3_DCTL_INITU2ENA)
 				usb_status |= 1 << USB_DEV_STAT_U2_ENABLED;
+		} else {
+			usb_status |= dwc->gadget->wakeup_armed <<
+					USB_DEVICE_REMOTE_WAKEUP;
 		}
 
 		break;
@@ -476,6 +479,10 @@ static int dwc3_ep0_handle_device(struct dwc3 *dwc,
 
 	switch (wValue) {
 	case USB_DEVICE_REMOTE_WAKEUP:
+		if (dwc->wakeup_configured)
+			dwc->gadget->wakeup_armed = set;
+		else
+			ret = -EINVAL;
 		break;
 	/*
 	 * 9.4.1 says only for SS, in AddressState only for
diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
index a1ebb30..c322c70 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
@@ -258,7 +258,7 @@ int dwc3_send_gadget_generic_command(struct dwc3 *dwc, unsigned int cmd,
 	return ret;
 }
 
-static int __dwc3_gadget_wakeup(struct dwc3 *dwc);
+static int __dwc3_gadget_wakeup(struct dwc3 *dwc, bool async);
 
 /**
  * dwc3_send_gadget_ep_cmd - issue an endpoint command
@@ -325,7 +325,7 @@ int dwc3_send_gadget_ep_cmd(struct dwc3_ep *dep, unsigned int cmd,
 
 			fallthrough;
 		case DWC3_LINK_STATE_U3:
-			ret = __dwc3_gadget_wakeup(dwc);
+			ret = __dwc3_gadget_wakeup(dwc, false);
 			dev_WARN_ONCE(dwc->dev, ret, "wakeup failed --> %d\n",
 					ret);
 			break;
@@ -2269,6 +2269,22 @@ static const struct usb_ep_ops dwc3_gadget_ep_ops = {
 
 /* -------------------------------------------------------------------------- */
 
+static void dwc3_gadget_enable_linksts_evts(struct dwc3 *dwc, bool set)
+{
+	u32 reg;
+
+	if (DWC3_VER_IS_PRIOR(DWC3, 250A))
+		return;
+
+	reg = dwc3_readl(dwc->regs, DWC3_DEVTEN);
+	if (set)
+		reg |= DWC3_DEVTEN_ULSTCNGEN;
+	else
+		reg &= ~DWC3_DEVTEN_ULSTCNGEN;
+
+	dwc3_writel(dwc->regs, DWC3_DEVTEN, reg);
+}
+
 static int dwc3_gadget_get_frame(struct usb_gadget *g)
 {
 	struct dwc3		*dwc = gadget_to_dwc(g);
@@ -2276,7 +2292,7 @@ static int dwc3_gadget_get_frame(struct usb_gadget *g)
 	return __dwc3_gadget_get_frame(dwc);
 }
 
-static int __dwc3_gadget_wakeup(struct dwc3 *dwc)
+static int __dwc3_gadget_wakeup(struct dwc3 *dwc, bool async)
 {
 	int			retries;
 
@@ -2307,9 +2323,13 @@ static int __dwc3_gadget_wakeup(struct dwc3 *dwc)
 		return -EINVAL;
 	}
 
+	if (async)
+		dwc3_gadget_enable_linksts_evts(dwc, true);
+
 	ret = dwc3_gadget_set_link_state(dwc, DWC3_LINK_STATE_RECOV);
 	if (ret < 0) {
 		dev_err(dwc->dev, "failed to put link in Recovery\n");
+		dwc3_gadget_enable_linksts_evts(dwc, false);
 		return ret;
 	}
 
@@ -2321,6 +2341,13 @@ static int __dwc3_gadget_wakeup(struct dwc3 *dwc)
 		dwc3_writel(dwc->regs, DWC3_DCTL, reg);
 	}
 
+	/*
+	 * Since link status change events are enabled we will receive
+	 * an U0 event when wakeup is successful. So bail out.
+	 */
+	if (async)
+		return 0;
+
 	/* poll until Link State changes to ON */
 	retries = 20000;
 
@@ -2346,13 +2373,36 @@ static int dwc3_gadget_wakeup(struct usb_gadget *g)
 	unsigned long		flags;
 	int			ret;
 
+	if (!dwc->wakeup_configured) {
+		dev_err(dwc->dev, "remote wakeup not configured\n");
+		return -EINVAL;
+	}
+
 	spin_lock_irqsave(&dwc->lock, flags);
-	ret = __dwc3_gadget_wakeup(dwc);
+	if (!dwc->gadget->wakeup_armed) {
+		dev_err(dwc->dev, "not armed for remote wakeup\n");
+		spin_unlock_irqrestore(&dwc->lock, flags);
+		return -EINVAL;
+	}
+	ret = __dwc3_gadget_wakeup(dwc, true);
+
 	spin_unlock_irqrestore(&dwc->lock, flags);
 
 	return ret;
 }
 
+static int dwc3_gadget_set_remote_wakeup(struct usb_gadget *g, int set)
+{
+	struct dwc3		*dwc = gadget_to_dwc(g);
+	unsigned long		flags;
+
+	spin_lock_irqsave(&dwc->lock, flags);
+	dwc->wakeup_configured = !!set;
+	spin_unlock_irqrestore(&dwc->lock, flags);
+
+	return 0;
+}
+
 static int dwc3_gadget_set_selfpowered(struct usb_gadget *g,
 		int is_selfpowered)
 {
@@ -2978,6 +3028,7 @@ static void dwc3_gadget_async_callbacks(struct usb_gadget *g, bool enable)
 static const struct usb_gadget_ops dwc3_gadget_ops = {
 	.get_frame		= dwc3_gadget_get_frame,
 	.wakeup			= dwc3_gadget_wakeup,
+	.set_remote_wakeup	= dwc3_gadget_set_remote_wakeup,
 	.set_selfpowered	= dwc3_gadget_set_selfpowered,
 	.pullup			= dwc3_gadget_pullup,
 	.udc_start		= dwc3_gadget_start,
@@ -3819,6 +3870,8 @@ static void dwc3_gadget_disconnect_interrupt(struct dwc3 *dwc)
 
 	dwc->gadget->speed = USB_SPEED_UNKNOWN;
 	dwc->setup_packet_pending = false;
+	dwc->gadget->wakeup_armed = false;
+	dwc3_gadget_enable_linksts_evts(dwc, false);
 	usb_gadget_set_state(dwc->gadget, USB_STATE_NOTATTACHED);
 
 	if (dwc->ep0state != EP0_SETUP_PHASE) {
@@ -3912,6 +3965,8 @@ static void dwc3_gadget_reset_interrupt(struct dwc3 *dwc)
 	reg &= ~DWC3_DCTL_TSTCTRL_MASK;
 	dwc3_gadget_dctl_write_safe(dwc, reg);
 	dwc->test_mode = false;
+	dwc->gadget->wakeup_armed = false;
+	dwc3_gadget_enable_linksts_evts(dwc, false);
 	dwc3_clear_stall_all_ep(dwc);
 
 	/* Reset device address to zero */
@@ -4064,7 +4119,7 @@ static void dwc3_gadget_conndone_interrupt(struct dwc3 *dwc)
 	 */
 }
 
-static void dwc3_gadget_wakeup_interrupt(struct dwc3 *dwc)
+static void dwc3_gadget_wakeup_interrupt(struct dwc3 *dwc, unsigned int evtinfo)
 {
 	/*
 	 * TODO take core out of low power mode when that's
@@ -4076,6 +4131,8 @@ static void dwc3_gadget_wakeup_interrupt(struct dwc3 *dwc)
 		dwc->gadget_driver->resume(dwc->gadget);
 		spin_lock(&dwc->lock);
 	}
+
+	dwc->link_state = evtinfo & DWC3_LINK_STATE_MASK;
 }
 
 static void dwc3_gadget_linksts_change_interrupt(struct dwc3 *dwc,
@@ -4157,6 +4214,12 @@ static void dwc3_gadget_linksts_change_interrupt(struct dwc3 *dwc,
 	}
 
 	switch (next) {
+	case DWC3_LINK_STATE_U0:
+		if (dwc->gadget->wakeup_armed) {
+			dwc3_gadget_enable_linksts_evts(dwc, false);
+			dwc3_resume_gadget(dwc);
+		}
+		break;
 	case DWC3_LINK_STATE_U1:
 		if (dwc->speed == USB_SPEED_SUPER)
 			dwc3_suspend_gadget(dwc);
@@ -4225,7 +4288,7 @@ static void dwc3_gadget_interrupt(struct dwc3 *dwc,
 		dwc3_gadget_conndone_interrupt(dwc);
 		break;
 	case DWC3_DEVICE_EVENT_WAKEUP:
-		dwc3_gadget_wakeup_interrupt(dwc);
+		dwc3_gadget_wakeup_interrupt(dwc, event->event_info);
 		break;
 	case DWC3_DEVICE_EVENT_HIBER_REQ:
 		if (dev_WARN_ONCE(dwc->dev, !dwc->has_hibernation,
@@ -4473,6 +4536,7 @@ int dwc3_gadget_init(struct dwc3 *dwc)
 	dwc->gadget->sg_supported	= true;
 	dwc->gadget->name		= "dwc3-gadget";
 	dwc->gadget->lpm_capable	= !dwc->usb2_gadget_lpm_disable;
+	dwc->gadget->wakeup_capable	= true;
 
 	/*
 	 * FIXME We might be setting max_speed to <SUPER, however versions
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 18+ messages in thread

* [PATCH v12 3/6] usb: gadget: Add function wakeup support
  2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 1/6] usb: gadget: Properly configure the device for remote wakeup Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 2/6] usb: dwc3: Add remote wakeup handling Elson Roy Serrao
@ 2023-03-16 23:38 ` Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 4/6] usb: dwc3: Add function suspend and " Elson Roy Serrao
                   ` (2 subsequent siblings)
  5 siblings, 0 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

USB3.2 spec section 9.2.5.4 quotes that a function may signal that
it wants to exit from Function Suspend by sending a Function
Wake Notification to the host if it is enabled for function
remote wakeup. Add an api in composite layer that can be used
by the function drivers to support this feature. Also expose
a gadget op so that composite layer can trigger a wakeup request
to the UDC driver.

Reviewed-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
---
 drivers/usb/gadget/composite.c | 40 ++++++++++++++++++++++++++++++++++++++++
 include/linux/usb/composite.h  |  6 ++++++
 include/linux/usb/gadget.h     |  1 +
 3 files changed, 47 insertions(+)

diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
index c9c983e..2111732 100644
--- a/drivers/usb/gadget/composite.c
+++ b/drivers/usb/gadget/composite.c
@@ -492,6 +492,46 @@ int usb_interface_id(struct usb_configuration *config,
 }
 EXPORT_SYMBOL_GPL(usb_interface_id);
 
+/**
+ * usb_func_wakeup - sends function wake notification to the host.
+ * @func: function that sends the remote wakeup notification.
+ *
+ * Applicable to devices operating at enhanced superspeed when usb
+ * functions are put in function suspend state and armed for function
+ * remote wakeup. On completion, function wake notification is sent. If
+ * the device is in low power state it tries to bring the device to active
+ * state before sending the wake notification. Since it is a synchronous
+ * call, caller must take care of not calling it in interrupt context.
+ * For devices operating at lower speeds  returns negative errno.
+ *
+ * Returns zero on success, else negative errno.
+ */
+int usb_func_wakeup(struct usb_function *func)
+{
+	struct usb_gadget	*gadget = func->config->cdev->gadget;
+	int			id;
+
+	if (!gadget->ops->func_wakeup)
+		return -EOPNOTSUPP;
+
+	if (!func->func_wakeup_armed) {
+		ERROR(func->config->cdev, "not armed for func remote wakeup\n");
+		return -EINVAL;
+	}
+
+	for (id = 0; id < MAX_CONFIG_INTERFACES; id++)
+		if (func->config->interface[id] == func)
+			break;
+
+	if (id == MAX_CONFIG_INTERFACES) {
+		ERROR(func->config->cdev, "Invalid function\n");
+		return -EINVAL;
+	}
+
+	return gadget->ops->func_wakeup(gadget, id);
+}
+EXPORT_SYMBOL_GPL(usb_func_wakeup);
+
 static u8 encode_bMaxPower(enum usb_device_speed speed,
 		struct usb_configuration *c)
 {
diff --git a/include/linux/usb/composite.h b/include/linux/usb/composite.h
index d949e91..a2448e9 100644
--- a/include/linux/usb/composite.h
+++ b/include/linux/usb/composite.h
@@ -150,6 +150,9 @@ struct usb_os_desc_table {
  *	GetStatus() request when the recipient is Interface.
  * @func_suspend: callback to be called when
  *	SetFeature(FUNCTION_SUSPEND) is reseived
+ * @func_suspended: Indicates whether the function is in function suspend state.
+ * @func_wakeup_armed: Indicates whether the function is armed by the host for
+ *	wakeup signaling.
  *
  * A single USB function uses one or more interfaces, and should in most
  * cases support operation at both full and high speeds.  Each function is
@@ -220,6 +223,8 @@ struct usb_function {
 	int			(*get_status)(struct usb_function *);
 	int			(*func_suspend)(struct usb_function *,
 						u8 suspend_opt);
+	bool			func_suspended;
+	bool			func_wakeup_armed;
 	/* private: */
 	/* internals */
 	struct list_head		list;
@@ -241,6 +246,7 @@ int config_ep_by_speed_and_alt(struct usb_gadget *g, struct usb_function *f,
 
 int config_ep_by_speed(struct usb_gadget *g, struct usb_function *f,
 			struct usb_ep *_ep);
+int usb_func_wakeup(struct usb_function *func);
 
 #define	MAX_CONFIG_INTERFACES		16	/* arbitrary; max 255 */
 
diff --git a/include/linux/usb/gadget.h b/include/linux/usb/gadget.h
index 1d79612..75bda078 100644
--- a/include/linux/usb/gadget.h
+++ b/include/linux/usb/gadget.h
@@ -310,6 +310,7 @@ struct usb_udc;
 struct usb_gadget_ops {
 	int	(*get_frame)(struct usb_gadget *);
 	int	(*wakeup)(struct usb_gadget *);
+	int	(*func_wakeup)(struct usb_gadget *gadget, int intf_id);
 	int	(*set_remote_wakeup)(struct usb_gadget *, int set);
 	int	(*set_selfpowered) (struct usb_gadget *, int is_selfpowered);
 	int	(*vbus_session) (struct usb_gadget *, int is_active);
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 18+ messages in thread

* [PATCH v12 4/6] usb: dwc3: Add function suspend and function wakeup support
  2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
                   ` (2 preceding siblings ...)
  2023-03-16 23:38 ` [PATCH v12 3/6] usb: gadget: Add function wakeup support Elson Roy Serrao
@ 2023-03-16 23:38 ` Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 5/6] usb: gadget: arm the function for triggering remote wakeup Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support Elson Roy Serrao
  5 siblings, 0 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

USB host sends function suspend and function resume notifications to
the interface through SET_FEATURE/CLEAR_FEATURE setup packets.
Add support to handle these packets by delegating the requests to
composite layer. Also add support to handle function wake notification
requests to exit from function suspend state.

Acked-by: Thinh Nguyen <Thinh.Nguyen@synopsys.com>
Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
---
 drivers/usb/dwc3/core.h   |  3 +++
 drivers/usb/dwc3/debug.h  |  2 ++
 drivers/usb/dwc3/ep0.c    | 12 ++++--------
 drivers/usb/dwc3/gadget.c | 42 ++++++++++++++++++++++++++++++++++++++++++
 4 files changed, 51 insertions(+), 8 deletions(-)

diff --git a/drivers/usb/dwc3/core.h b/drivers/usb/dwc3/core.h
index 416e0ef..3412ebb 100644
--- a/drivers/usb/dwc3/core.h
+++ b/drivers/usb/dwc3/core.h
@@ -527,6 +527,7 @@
 #define DWC3_DGCMD_SET_ENDPOINT_NRDY	0x0c
 #define DWC3_DGCMD_SET_ENDPOINT_PRIME	0x0d
 #define DWC3_DGCMD_RUN_SOC_BUS_LOOPBACK	0x10
+#define DWC3_DGCMD_DEV_NOTIFICATION	0x07
 
 #define DWC3_DGCMD_STATUS(n)		(((n) >> 12) & 0x0F)
 #define DWC3_DGCMD_CMDACT		BIT(10)
@@ -539,6 +540,8 @@
 #define DWC3_DGCMDPAR_TX_FIFO			BIT(5)
 #define DWC3_DGCMDPAR_LOOPBACK_DIS		(0 << 0)
 #define DWC3_DGCMDPAR_LOOPBACK_ENA		BIT(0)
+#define DWC3_DGCMDPAR_DN_FUNC_WAKE		BIT(0)
+#define DWC3_DGCMDPAR_INTF_SEL(n)		((n) << 4)
 
 /* Device Endpoint Command Register */
 #define DWC3_DEPCMD_PARAM_SHIFT		16
diff --git a/drivers/usb/dwc3/debug.h b/drivers/usb/dwc3/debug.h
index 8bb2c9e..09d7038 100644
--- a/drivers/usb/dwc3/debug.h
+++ b/drivers/usb/dwc3/debug.h
@@ -72,6 +72,8 @@ dwc3_gadget_generic_cmd_string(u8 cmd)
 		return "Set Endpoint Prime";
 	case DWC3_DGCMD_RUN_SOC_BUS_LOOPBACK:
 		return "Run SoC Bus Loopback Test";
+	case DWC3_DGCMD_DEV_NOTIFICATION:
+		return "Device Notification";
 	default:
 		return "UNKNOWN";
 	}
diff --git a/drivers/usb/dwc3/ep0.c b/drivers/usb/dwc3/ep0.c
index 8aa14a5..953b752 100644
--- a/drivers/usb/dwc3/ep0.c
+++ b/drivers/usb/dwc3/ep0.c
@@ -30,6 +30,8 @@
 static void __dwc3_ep0_do_control_status(struct dwc3 *dwc, struct dwc3_ep *dep);
 static void __dwc3_ep0_do_control_data(struct dwc3 *dwc,
 		struct dwc3_ep *dep, struct dwc3_request *req);
+static int dwc3_ep0_delegate_req(struct dwc3 *dwc,
+				 struct usb_ctrlrequest *ctrl);
 
 static void dwc3_ep0_prepare_one_trb(struct dwc3_ep *dep,
 		dma_addr_t buf_dma, u32 len, u32 type, bool chain)
@@ -368,7 +370,7 @@ static int dwc3_ep0_handle_status(struct dwc3 *dwc,
 		 * Function Remote Wake Capable	D0
 		 * Function Remote Wakeup	D1
 		 */
-		break;
+		return dwc3_ep0_delegate_req(dwc, ctrl);
 
 	case USB_RECIP_ENDPOINT:
 		dep = dwc3_wIndex_to_dep(dwc, ctrl->wIndex);
@@ -517,13 +519,7 @@ static int dwc3_ep0_handle_intf(struct dwc3 *dwc,
 
 	switch (wValue) {
 	case USB_INTRF_FUNC_SUSPEND:
-		/*
-		 * REVISIT: Ideally we would enable some low power mode here,
-		 * however it's unclear what we should be doing here.
-		 *
-		 * For now, we're not doing anything, just making sure we return
-		 * 0 so USB Command Verifier tests pass without any errors.
-		 */
+		ret = dwc3_ep0_delegate_req(dwc, ctrl);
 		break;
 	default:
 		ret = -EINVAL;
diff --git a/drivers/usb/dwc3/gadget.c b/drivers/usb/dwc3/gadget.c
index c322c70..9c3d619 100644
--- a/drivers/usb/dwc3/gadget.c
+++ b/drivers/usb/dwc3/gadget.c
@@ -2391,6 +2391,47 @@ static int dwc3_gadget_wakeup(struct usb_gadget *g)
 	return ret;
 }
 
+static void dwc3_resume_gadget(struct dwc3 *dwc);
+
+static int dwc3_gadget_func_wakeup(struct usb_gadget *g, int intf_id)
+{
+	struct  dwc3		*dwc = gadget_to_dwc(g);
+	unsigned long		flags;
+	int			ret;
+	int			link_state;
+
+	if (!dwc->wakeup_configured) {
+		dev_err(dwc->dev, "remote wakeup not configured\n");
+		return -EINVAL;
+	}
+
+	spin_lock_irqsave(&dwc->lock, flags);
+	/*
+	 * If the link is in U3, signal for remote wakeup and wait for the
+	 * link to transition to U0 before sending device notification.
+	 */
+	link_state = dwc3_gadget_get_link_state(dwc);
+	if (link_state == DWC3_LINK_STATE_U3) {
+		ret = __dwc3_gadget_wakeup(dwc, false);
+		if (ret) {
+			spin_unlock_irqrestore(&dwc->lock, flags);
+			return -EINVAL;
+		}
+		dwc3_resume_gadget(dwc);
+		dwc->link_state = DWC3_LINK_STATE_U0;
+	}
+
+	ret = dwc3_send_gadget_generic_command(dwc, DWC3_DGCMD_DEV_NOTIFICATION,
+					       DWC3_DGCMDPAR_DN_FUNC_WAKE |
+					       DWC3_DGCMDPAR_INTF_SEL(intf_id));
+	if (ret)
+		dev_err(dwc->dev, "function remote wakeup failed, ret:%d\n", ret);
+
+	spin_unlock_irqrestore(&dwc->lock, flags);
+
+	return ret;
+}
+
 static int dwc3_gadget_set_remote_wakeup(struct usb_gadget *g, int set)
 {
 	struct dwc3		*dwc = gadget_to_dwc(g);
@@ -3028,6 +3069,7 @@ static void dwc3_gadget_async_callbacks(struct usb_gadget *g, bool enable)
 static const struct usb_gadget_ops dwc3_gadget_ops = {
 	.get_frame		= dwc3_gadget_get_frame,
 	.wakeup			= dwc3_gadget_wakeup,
+	.func_wakeup		= dwc3_gadget_func_wakeup,
 	.set_remote_wakeup	= dwc3_gadget_set_remote_wakeup,
 	.set_selfpowered	= dwc3_gadget_set_selfpowered,
 	.pullup			= dwc3_gadget_pullup,
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 18+ messages in thread

* [PATCH v12 5/6] usb: gadget: arm the function for triggering remote wakeup
  2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
                   ` (3 preceding siblings ...)
  2023-03-16 23:38 ` [PATCH v12 4/6] usb: dwc3: Add function suspend and " Elson Roy Serrao
@ 2023-03-16 23:38 ` Elson Roy Serrao
  2023-03-16 23:38 ` [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support Elson Roy Serrao
  5 siblings, 0 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

When host sends function suspend feature selector to the device,
arm the function for remote wakeup based on the received packet.
Also host queries the function wakeup capability  through a get
status request. Handle such requests in composite layer for cases
where function driver has not exposed a get_status callback.

Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
---
 drivers/usb/gadget/composite.c | 26 +++++++++++++++++++++++---
 1 file changed, 23 insertions(+), 3 deletions(-)

diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
index 2111732..2c1100e 100644
--- a/drivers/usb/gadget/composite.c
+++ b/drivers/usb/gadget/composite.c
@@ -2006,9 +2006,20 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
 		f = cdev->config->interface[intf];
 		if (!f)
 			break;
-		status = f->get_status ? f->get_status(f) : 0;
-		if (status < 0)
-			break;
+
+		if (f->get_status) {
+			status = f->get_status(f);
+			if (status < 0)
+				break;
+		} else {
+			/* Set D0 and D1 bits based on func wakeup capability */
+			if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
+				status |= USB_INTRF_STAT_FUNC_RW_CAP;
+				if (f->func_wakeup_armed)
+					status |= USB_INTRF_STAT_FUNC_RW;
+			}
+		}
+
 		put_unaligned_le16(status & 0x0000ffff, req->buf);
 		break;
 	/*
@@ -2029,6 +2040,15 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
 			f = cdev->config->interface[intf];
 			if (!f)
 				break;
+
+			if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
+				if (!(f->config->bmAttributes &
+				      USB_CONFIG_ATT_WAKEUP))
+					break;
+				f->func_wakeup_armed = (ctrl->bRequest ==
+							USB_REQ_SET_FEATURE);
+			}
+
 			value = 0;
 			if (f->func_suspend)
 				value = f->func_suspend(f, w_index >> 8);
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 18+ messages in thread

* [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
                   ` (4 preceding siblings ...)
  2023-03-16 23:38 ` [PATCH v12 5/6] usb: gadget: arm the function for triggering remote wakeup Elson Roy Serrao
@ 2023-03-16 23:38 ` Elson Roy Serrao
  2023-03-17  0:11   ` Thinh Nguyen
  2023-03-21  8:12   ` Dan Carpenter
  5 siblings, 2 replies; 18+ messages in thread
From: Elson Roy Serrao @ 2023-03-16 23:38 UTC (permalink / raw)
  To: gregkh, Thinh.Nguyen
  Cc: linux-kernel, linux-usb, quic_wcheng, quic_jackp, Elson Roy Serrao

When host sends a suspend notification to the device, handle
the suspend callbacks in the function driver. Enhanced super
speed devices can support function suspend feature to put the
function in suspend state. Handle function suspend callback.

Depending on the remote wakeup capability the device can either
trigger a remote wakeup or wait for the host initiated resume to
start data transfer again.

Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
---
 drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
 drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
 drivers/usb/gadget/function/u_ether.h |  4 ++
 3 files changed, 138 insertions(+)

diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
index a7ab30e..c43cd557 100644
--- a/drivers/usb/gadget/function/f_ecm.c
+++ b/drivers/usb/gadget/function/f_ecm.c
@@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
 
 	usb_ep_disable(ecm->notify);
 	ecm->notify->desc = NULL;
+	f->func_suspended = false;
+	f->func_wakeup_armed = false;
 }
 
 /*-------------------------------------------------------------------------*/
@@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
 	return &opts->func_inst;
 }
 
+static void ecm_suspend(struct usb_function *f)
+{
+	struct f_ecm *ecm = func_to_ecm(f);
+	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
+
+	if (f->func_suspended) {
+		DBG(cdev, "Function already suspended\n");
+		return;
+	}
+
+	DBG(cdev, "ECM Suspend\n");
+
+	gether_suspend(&ecm->port);
+}
+
+static void ecm_resume(struct usb_function *f)
+{
+	struct f_ecm *ecm = func_to_ecm(f);
+	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
+
+	/*
+	 * If the function is in USB3 Function Suspend state, resume is
+	 * canceled. In this case resume is done by a Function Resume request.
+	 */
+	if (f->func_suspended)
+		return;
+
+	DBG(cdev, "ECM Resume\n");
+
+	gether_resume(&ecm->port);
+}
+
+static int ecm_get_status(struct usb_function *f)
+{
+	struct usb_configuration *c = f->config;
+
+	/* D0 and D1 bit set to 0 if device is not wakeup capable */
+	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
+		return 0;
+
+	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
+		USB_INTRF_STAT_FUNC_RW_CAP;
+}
+
+static int ecm_func_suspend(struct usb_function *f, u8 options)
+{
+	struct usb_composite_dev *cdev = f->config->cdev;
+
+	DBG(cdev, "func susp %u cmd\n", options);
+
+	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
+		if (!f->func_suspended) {
+			ecm_suspend(f);
+			f->func_suspended = true;
+		}
+	} else {
+		if (f->func_suspended) {
+			f->func_suspended = false;
+			ecm_resume(f);
+		}
+	}
+
+	return 0;
+}
+
 static void ecm_free(struct usb_function *f)
 {
 	struct f_ecm *ecm;
@@ -952,6 +1019,10 @@ static struct usb_function *ecm_alloc(struct usb_function_instance *fi)
 	ecm->port.func.setup = ecm_setup;
 	ecm->port.func.disable = ecm_disable;
 	ecm->port.func.free_func = ecm_free;
+	ecm->port.func.suspend = ecm_suspend;
+	ecm->port.func.get_status = ecm_get_status;
+	ecm->port.func.func_suspend = ecm_func_suspend;
+	ecm->port.func.resume = ecm_resume;
 
 	return &ecm->port.func;
 }
diff --git a/drivers/usb/gadget/function/u_ether.c b/drivers/usb/gadget/function/u_ether.c
index f259975..8eba018 100644
--- a/drivers/usb/gadget/function/u_ether.c
+++ b/drivers/usb/gadget/function/u_ether.c
@@ -437,6 +437,20 @@ static inline int is_promisc(u16 cdc_filter)
 	return cdc_filter & USB_CDC_PACKET_TYPE_PROMISCUOUS;
 }
 
+static int ether_wakeup_host(struct gether *port)
+{
+	int			ret;
+	struct usb_function	*func = &port->func;
+	struct usb_gadget	*gadget = func->config->cdev->gadget;
+
+	if (func->func_suspended)
+		ret = usb_func_wakeup(func);
+	else
+		ret = usb_gadget_wakeup(gadget);
+
+	return ret;
+}
+
 static netdev_tx_t eth_start_xmit(struct sk_buff *skb,
 					struct net_device *net)
 {
@@ -456,6 +470,15 @@ static netdev_tx_t eth_start_xmit(struct sk_buff *skb,
 		in = NULL;
 		cdc_filter = 0;
 	}
+
+	if (dev->port_usb->is_suspend) {
+		DBG(dev, "Port suspended. Triggering wakeup\n");
+		netif_stop_queue(net);
+		spin_unlock_irqrestore(&dev->lock, flags);
+		ether_wakeup_host(dev->port_usb);
+		return NETDEV_TX_BUSY;
+	}
+
 	spin_unlock_irqrestore(&dev->lock, flags);
 
 	if (!in) {
@@ -1014,6 +1037,45 @@ int gether_set_ifname(struct net_device *net, const char *name, int len)
 }
 EXPORT_SYMBOL_GPL(gether_set_ifname);
 
+void gether_suspend(struct gether *link)
+{
+	struct eth_dev *dev = link->ioport;
+	unsigned long flags;
+
+	if (!dev)
+		return;
+
+	if (atomic_read(&dev->tx_qlen)) {
+		/*
+		 * There is a transfer in progress. So we trigger a remote
+		 * wakeup to inform the host.
+		 */
+		ether_wakeup_host(dev->port_usb);
+		return;
+	}
+	spin_lock_irqsave(&dev->lock, flags);
+	link->is_suspend = true;
+	spin_unlock_irqrestore(&dev->lock, flags);
+}
+EXPORT_SYMBOL_GPL(gether_suspend);
+
+void gether_resume(struct gether *link)
+{
+	struct eth_dev *dev = link->ioport;
+	unsigned long flags;
+
+	if (!dev)
+		return;
+
+	if (netif_queue_stopped(dev->net))
+		netif_start_queue(dev->net);
+
+	spin_lock_irqsave(&dev->lock, flags);
+	link->is_suspend = false;
+	spin_unlock_irqrestore(&dev->lock, flags);
+}
+EXPORT_SYMBOL_GPL(gether_resume);
+
 /*
  * gether_cleanup - remove Ethernet-over-USB device
  * Context: may sleep
@@ -1176,6 +1238,7 @@ void gether_disconnect(struct gether *link)
 
 	spin_lock(&dev->lock);
 	dev->port_usb = NULL;
+	link->is_suspend = false;
 	spin_unlock(&dev->lock);
 }
 EXPORT_SYMBOL_GPL(gether_disconnect);
diff --git a/drivers/usb/gadget/function/u_ether.h b/drivers/usb/gadget/function/u_ether.h
index 4014454..851ee10 100644
--- a/drivers/usb/gadget/function/u_ether.h
+++ b/drivers/usb/gadget/function/u_ether.h
@@ -79,6 +79,7 @@ struct gether {
 	/* called on network open/close */
 	void				(*open)(struct gether *);
 	void				(*close)(struct gether *);
+	bool				is_suspend;
 };
 
 #define	DEFAULT_FILTER	(USB_CDC_PACKET_TYPE_BROADCAST \
@@ -258,6 +259,9 @@ int gether_set_ifname(struct net_device *net, const char *name, int len);
 
 void gether_cleanup(struct eth_dev *dev);
 
+void gether_suspend(struct gether *link);
+void gether_resume(struct gether *link);
+
 /* connect/disconnect is handled by individual functions */
 struct net_device *gether_connect(struct gether *);
 void gether_disconnect(struct gether *);
-- 
2.7.4


^ permalink raw reply related	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-16 23:38 ` [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support Elson Roy Serrao
@ 2023-03-17  0:11   ` Thinh Nguyen
  2023-03-17 17:59     ` Elson Serrao
  2023-03-21  8:12   ` Dan Carpenter
  1 sibling, 1 reply; 18+ messages in thread
From: Thinh Nguyen @ 2023-03-17  0:11 UTC (permalink / raw)
  To: Elson Roy Serrao
  Cc: gregkh, Thinh Nguyen, linux-kernel, linux-usb, quic_wcheng, quic_jackp

On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
> When host sends a suspend notification to the device, handle
> the suspend callbacks in the function driver. Enhanced super
> speed devices can support function suspend feature to put the
> function in suspend state. Handle function suspend callback.
> 
> Depending on the remote wakeup capability the device can either
> trigger a remote wakeup or wait for the host initiated resume to
> start data transfer again.
> 
> Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
> ---
>  drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
>  drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
>  drivers/usb/gadget/function/u_ether.h |  4 ++
>  3 files changed, 138 insertions(+)
> 
> diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
> index a7ab30e..c43cd557 100644
> --- a/drivers/usb/gadget/function/f_ecm.c
> +++ b/drivers/usb/gadget/function/f_ecm.c
> @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
>  
>  	usb_ep_disable(ecm->notify);
>  	ecm->notify->desc = NULL;
> +	f->func_suspended = false;
> +	f->func_wakeup_armed = false;
>  }
>  
>  /*-------------------------------------------------------------------------*/
> @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
>  	return &opts->func_inst;
>  }
>  
> +static void ecm_suspend(struct usb_function *f)
> +{
> +	struct f_ecm *ecm = func_to_ecm(f);
> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> +
> +	if (f->func_suspended) {
> +		DBG(cdev, "Function already suspended\n");
> +		return;
> +	}
> +
> +	DBG(cdev, "ECM Suspend\n");
> +
> +	gether_suspend(&ecm->port);
> +}
> +
> +static void ecm_resume(struct usb_function *f)
> +{
> +	struct f_ecm *ecm = func_to_ecm(f);
> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> +
> +	/*
> +	 * If the function is in USB3 Function Suspend state, resume is
> +	 * canceled. In this case resume is done by a Function Resume request.
> +	 */
> +	if (f->func_suspended)
> +		return;
> +
> +	DBG(cdev, "ECM Resume\n");
> +
> +	gether_resume(&ecm->port);
> +}
> +
> +static int ecm_get_status(struct usb_function *f)
> +{
> +	struct usb_configuration *c = f->config;
> +
> +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
> +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
> +		return 0;
> +
> +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
> +		USB_INTRF_STAT_FUNC_RW_CAP;
> +}

Why do we need to implement ecm_get_status if it's already handled in
composite.c now?

> +
> +static int ecm_func_suspend(struct usb_function *f, u8 options)
> +{
> +	struct usb_composite_dev *cdev = f->config->cdev;
> +
> +	DBG(cdev, "func susp %u cmd\n", options);
> +
> +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {

This feature selector doesn't indicate whether it's SetFeature or
ClearFeature request. ecm_func_suspend is supposed to be for
SetFeature(suspend) only. Perhaps we may have to define func_resume()
for ClearFeature(suspend)?

Thanks,
Thinh

> +		if (!f->func_suspended) {
> +			ecm_suspend(f);
> +			f->func_suspended = true;
> +		}
> +	} else {
> +		if (f->func_suspended) {
> +			f->func_suspended = false;
> +			ecm_resume(f);
> +		}
> +	}
> +
> +	return 0;
> +}
> +

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-17  0:11   ` Thinh Nguyen
@ 2023-03-17 17:59     ` Elson Serrao
  2023-03-17 21:28       ` Thinh Nguyen
  0 siblings, 1 reply; 18+ messages in thread
From: Elson Serrao @ 2023-03-17 17:59 UTC (permalink / raw)
  To: Thinh Nguyen; +Cc: gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp



On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
> On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
>> When host sends a suspend notification to the device, handle
>> the suspend callbacks in the function driver. Enhanced super
>> speed devices can support function suspend feature to put the
>> function in suspend state. Handle function suspend callback.
>>
>> Depending on the remote wakeup capability the device can either
>> trigger a remote wakeup or wait for the host initiated resume to
>> start data transfer again.
>>
>> Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
>> ---
>>   drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
>>   drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
>>   drivers/usb/gadget/function/u_ether.h |  4 ++
>>   3 files changed, 138 insertions(+)
>>
>> diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
>> index a7ab30e..c43cd557 100644
>> --- a/drivers/usb/gadget/function/f_ecm.c
>> +++ b/drivers/usb/gadget/function/f_ecm.c
>> @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
>>   
>>   	usb_ep_disable(ecm->notify);
>>   	ecm->notify->desc = NULL;
>> +	f->func_suspended = false;
>> +	f->func_wakeup_armed = false;
>>   }
>>   
>>   /*-------------------------------------------------------------------------*/
>> @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
>>   	return &opts->func_inst;
>>   }
>>   
>> +static void ecm_suspend(struct usb_function *f)
>> +{
>> +	struct f_ecm *ecm = func_to_ecm(f);
>> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>> +
>> +	if (f->func_suspended) {
>> +		DBG(cdev, "Function already suspended\n");
>> +		return;
>> +	}
>> +
>> +	DBG(cdev, "ECM Suspend\n");
>> +
>> +	gether_suspend(&ecm->port);
>> +}
>> +
>> +static void ecm_resume(struct usb_function *f)
>> +{
>> +	struct f_ecm *ecm = func_to_ecm(f);
>> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>> +
>> +	/*
>> +	 * If the function is in USB3 Function Suspend state, resume is
>> +	 * canceled. In this case resume is done by a Function Resume request.
>> +	 */
>> +	if (f->func_suspended)
>> +		return;
>> +
>> +	DBG(cdev, "ECM Resume\n");
>> +
>> +	gether_resume(&ecm->port);
>> +}
>> +
>> +static int ecm_get_status(struct usb_function *f)
>> +{
>> +	struct usb_configuration *c = f->config;
>> +
>> +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
>> +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
>> +		return 0;
>> +
>> +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
>> +		USB_INTRF_STAT_FUNC_RW_CAP;
>> +}
> 
> Why do we need to implement ecm_get_status if it's already handled in
> composite.c now?
> 

Yes this can be removed now. Will modify accordingly.
>> +
>> +static int ecm_func_suspend(struct usb_function *f, u8 options)
>> +{
>> +	struct usb_composite_dev *cdev = f->config->cdev;
>> +
>> +	DBG(cdev, "func susp %u cmd\n", options);
>> +
>> +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
> 
> This feature selector doesn't indicate whether it's SetFeature or
> ClearFeature request. ecm_func_suspend is supposed to be for
> SetFeature(suspend) only. Perhaps we may have to define func_resume()
> for ClearFeature(suspend)?
> 
> Thanks,
> Thinh
> 
Host uses the same feature selector FUNCTION_SUSPEND for function 
suspend and function resume and func_suspend() callback can be used to
handle both the cases ? The distinction comes whether it is a 
SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) which can 
be easily done in the func_suspend callback itself. We can add another 
callback func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont 
that be redundant and more callback handling on function 
driver/composite side as well? Please let me know your opinion.

Thanks
Elson

>> +		if (!f->func_suspended) {
>> +			ecm_suspend(f);
>> +			f->func_suspended = true;
>> +		}
>> +	} else {
>> +		if (f->func_suspended) {
>> +			f->func_suspended = false;
>> +			ecm_resume(f);
>> +		}
>> +	}
>> +
>> +	return 0;
>> +}
>> +

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-17 17:59     ` Elson Serrao
@ 2023-03-17 21:28       ` Thinh Nguyen
  2023-03-17 23:20         ` Thinh Nguyen
  0 siblings, 1 reply; 18+ messages in thread
From: Thinh Nguyen @ 2023-03-17 21:28 UTC (permalink / raw)
  To: Elson Serrao
  Cc: Thinh Nguyen, gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp

On Fri, Mar 17, 2023, Elson Serrao wrote:
> 
> 
> On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
> > On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
> > > When host sends a suspend notification to the device, handle
> > > the suspend callbacks in the function driver. Enhanced super
> > > speed devices can support function suspend feature to put the
> > > function in suspend state. Handle function suspend callback.
> > > 
> > > Depending on the remote wakeup capability the device can either
> > > trigger a remote wakeup or wait for the host initiated resume to
> > > start data transfer again.
> > > 
> > > Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
> > > ---
> > >   drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
> > >   drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
> > >   drivers/usb/gadget/function/u_ether.h |  4 ++
> > >   3 files changed, 138 insertions(+)
> > > 
> > > diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
> > > index a7ab30e..c43cd557 100644
> > > --- a/drivers/usb/gadget/function/f_ecm.c
> > > +++ b/drivers/usb/gadget/function/f_ecm.c
> > > @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
> > >   	usb_ep_disable(ecm->notify);
> > >   	ecm->notify->desc = NULL;
> > > +	f->func_suspended = false;
> > > +	f->func_wakeup_armed = false;
> > >   }
> > >   /*-------------------------------------------------------------------------*/
> > > @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
> > >   	return &opts->func_inst;
> > >   }
> > > +static void ecm_suspend(struct usb_function *f)
> > > +{
> > > +	struct f_ecm *ecm = func_to_ecm(f);
> > > +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > +
> > > +	if (f->func_suspended) {
> > > +		DBG(cdev, "Function already suspended\n");
> > > +		return;
> > > +	}
> > > +
> > > +	DBG(cdev, "ECM Suspend\n");
> > > +
> > > +	gether_suspend(&ecm->port);
> > > +}
> > > +
> > > +static void ecm_resume(struct usb_function *f)
> > > +{
> > > +	struct f_ecm *ecm = func_to_ecm(f);
> > > +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > +
> > > +	/*
> > > +	 * If the function is in USB3 Function Suspend state, resume is
> > > +	 * canceled. In this case resume is done by a Function Resume request.
> > > +	 */
> > > +	if (f->func_suspended)
> > > +		return;
> > > +
> > > +	DBG(cdev, "ECM Resume\n");
> > > +
> > > +	gether_resume(&ecm->port);
> > > +}
> > > +
> > > +static int ecm_get_status(struct usb_function *f)
> > > +{
> > > +	struct usb_configuration *c = f->config;
> > > +
> > > +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
> > > +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
> > > +		return 0;
> > > +
> > > +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
> > > +		USB_INTRF_STAT_FUNC_RW_CAP;
> > > +}
> > 
> > Why do we need to implement ecm_get_status if it's already handled in
> > composite.c now?
> > 
> 
> Yes this can be removed now. Will modify accordingly.
> > > +
> > > +static int ecm_func_suspend(struct usb_function *f, u8 options)
> > > +{
> > > +	struct usb_composite_dev *cdev = f->config->cdev;
> > > +
> > > +	DBG(cdev, "func susp %u cmd\n", options);
> > > +
> > > +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
> > 
> > This feature selector doesn't indicate whether it's SetFeature or
> > ClearFeature request. ecm_func_suspend is supposed to be for
> > SetFeature(suspend) only. Perhaps we may have to define func_resume()
> > for ClearFeature(suspend)?
> > 

> Host uses the same feature selector FUNCTION_SUSPEND for function suspend
> and function resume and func_suspend() callback can be used to
> handle both the cases ? The distinction comes whether it is a

How do you plan to handle that? Pass this info in some unused/reserved
bit of the "options" argument? Introduce a new parameter to the
func_suspend()?

If that's the case, then you need to update the document on
func_suspend() to also support ClearFeature(suspend). Right now it's
documented for SetFeature only. Also, make sure that other existing
function drivers will not break because of the change of the
func_suspend behavior.

> SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) which can be
> easily done in the func_suspend callback itself. We can add another callback
> func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont that be
> redundant and more callback handling on function driver/composite side as
> well? Please let me know your opinion.
> 

We actually didn't properly define func_suspend and its counter part. It
seems cleaner to me to introduce func_resume as it seems more intuitive
and easier to read. Let me know how you plan to use func_suspend() for
both cases.

Thanks,
Thinh

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-17 21:28       ` Thinh Nguyen
@ 2023-03-17 23:20         ` Thinh Nguyen
  2023-03-18  0:23           ` Elson Serrao
  0 siblings, 1 reply; 18+ messages in thread
From: Thinh Nguyen @ 2023-03-17 23:20 UTC (permalink / raw)
  To: Elson Serrao; +Cc: gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp

On Fri, Mar 17, 2023, Thinh Nguyen wrote:
> On Fri, Mar 17, 2023, Elson Serrao wrote:
> > 
> > 
> > On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
> > > On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
> > > > When host sends a suspend notification to the device, handle
> > > > the suspend callbacks in the function driver. Enhanced super
> > > > speed devices can support function suspend feature to put the
> > > > function in suspend state. Handle function suspend callback.
> > > > 
> > > > Depending on the remote wakeup capability the device can either
> > > > trigger a remote wakeup or wait for the host initiated resume to
> > > > start data transfer again.
> > > > 
> > > > Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
> > > > ---
> > > >   drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
> > > >   drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
> > > >   drivers/usb/gadget/function/u_ether.h |  4 ++
> > > >   3 files changed, 138 insertions(+)
> > > > 
> > > > diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
> > > > index a7ab30e..c43cd557 100644
> > > > --- a/drivers/usb/gadget/function/f_ecm.c
> > > > +++ b/drivers/usb/gadget/function/f_ecm.c
> > > > @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
> > > >   	usb_ep_disable(ecm->notify);
> > > >   	ecm->notify->desc = NULL;
> > > > +	f->func_suspended = false;
> > > > +	f->func_wakeup_armed = false;
> > > >   }
> > > >   /*-------------------------------------------------------------------------*/
> > > > @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
> > > >   	return &opts->func_inst;
> > > >   }
> > > > +static void ecm_suspend(struct usb_function *f)
> > > > +{
> > > > +	struct f_ecm *ecm = func_to_ecm(f);
> > > > +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > +
> > > > +	if (f->func_suspended) {
> > > > +		DBG(cdev, "Function already suspended\n");
> > > > +		return;
> > > > +	}
> > > > +
> > > > +	DBG(cdev, "ECM Suspend\n");
> > > > +
> > > > +	gether_suspend(&ecm->port);
> > > > +}
> > > > +
> > > > +static void ecm_resume(struct usb_function *f)
> > > > +{
> > > > +	struct f_ecm *ecm = func_to_ecm(f);
> > > > +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > +
> > > > +	/*
> > > > +	 * If the function is in USB3 Function Suspend state, resume is
> > > > +	 * canceled. In this case resume is done by a Function Resume request.
> > > > +	 */
> > > > +	if (f->func_suspended)
> > > > +		return;
> > > > +
> > > > +	DBG(cdev, "ECM Resume\n");
> > > > +
> > > > +	gether_resume(&ecm->port);
> > > > +}
> > > > +
> > > > +static int ecm_get_status(struct usb_function *f)
> > > > +{
> > > > +	struct usb_configuration *c = f->config;
> > > > +
> > > > +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
> > > > +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
> > > > +		return 0;
> > > > +
> > > > +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
> > > > +		USB_INTRF_STAT_FUNC_RW_CAP;
> > > > +}
> > > 
> > > Why do we need to implement ecm_get_status if it's already handled in
> > > composite.c now?
> > > 
> > 
> > Yes this can be removed now. Will modify accordingly.
> > > > +
> > > > +static int ecm_func_suspend(struct usb_function *f, u8 options)
> > > > +{
> > > > +	struct usb_composite_dev *cdev = f->config->cdev;
> > > > +
> > > > +	DBG(cdev, "func susp %u cmd\n", options);
> > > > +
> > > > +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
> > > 
> > > This feature selector doesn't indicate whether it's SetFeature or
> > > ClearFeature request. ecm_func_suspend is supposed to be for
> > > SetFeature(suspend) only. Perhaps we may have to define func_resume()
> > > for ClearFeature(suspend)?
> > > 
> 
> > Host uses the same feature selector FUNCTION_SUSPEND for function suspend
> > and function resume and func_suspend() callback can be used to
> > handle both the cases ? The distinction comes whether it is a
> 
> How do you plan to handle that? Pass this info in some unused/reserved
> bit of the "options" argument? Introduce a new parameter to the
> func_suspend()?
> 
> If that's the case, then you need to update the document on
> func_suspend() to also support ClearFeature(suspend). Right now it's
> documented for SetFeature only. Also, make sure that other existing
> function drivers will not break because of the change of the
> func_suspend behavior.
> 
> > SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) which can be
> > easily done in the func_suspend callback itself. We can add another callback
> > func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont that be
> > redundant and more callback handling on function driver/composite side as
> > well? Please let me know your opinion.
> > 
> 
> We actually didn't properly define func_suspend and its counter part. It
> seems cleaner to me to introduce func_resume as it seems more intuitive
> and easier to read. Let me know how you plan to use func_suspend() for
> both cases.
> 

How about we handle function suspend resume in composite also? I mean
something like this:

diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
index 36add1879ed2..79dc055eb5f7 100644
--- a/drivers/usb/gadget/composite.c
+++ b/drivers/usb/gadget/composite.c
@@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
 		f = cdev->config->interface[intf];
 		if (!f)
 			break;
-		status = f->get_status ? f->get_status(f) : 0;
-		if (status < 0)
-			break;
+
+		if (f->get_status) {
+			status = f->get_status(f);
+			if (status < 0)
+				break;
+		} else {
+			if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
+				status |= USB_INTRF_STAT_FUNC_RW_CAP;
+				if (f->func_wakeup_armed)
+					status |= USB_INTRF_STAT_FUNC_RW;
+			}
+		}
 		put_unaligned_le16(status & 0x0000ffff, req->buf);
 		break;
 	/*
@@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
 			f = cdev->config->interface[intf];
 			if (!f)
 				break;
+			if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
+				if (!(f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP))
+					break;
+
+				f->func_wakeup_armed = (ctrl->bRequest == USB_REQ_SET_FEATURE);
+			}
+
 			value = 0;
-			if (f->func_suspend)
+			if (f->func_suspend) {
 				value = f->func_suspend(f, w_index >> 8);
+			} else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
+				if (f->suspend && && !f->func_suspended &&
+				    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
+					f->suspend(f);
+					f->func_suspended = true;
+				} else if (f->resume && f->func_suspended &&
+					   ctrl->bRequest == USB_REQ_CLEAR_FEATURE_FEATURE)) {
+					f->resume(f);
+					f->func_suspended = false;
+				}
+			}
+
 			if (value < 0) {
 				ERROR(cdev,
 				      "func_suspend() returned error %d\n",


Also, do we need the f->func_suspended flag? we'd need the remote wakeup
flag for the status, but when do we need f->func_suspended? It seems
like it can be handled within the function driver's scope.

Thanks,
Thinh

^ permalink raw reply related	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-17 23:20         ` Thinh Nguyen
@ 2023-03-18  0:23           ` Elson Serrao
  2023-03-18  2:26             ` Thinh Nguyen
  0 siblings, 1 reply; 18+ messages in thread
From: Elson Serrao @ 2023-03-18  0:23 UTC (permalink / raw)
  To: Thinh Nguyen; +Cc: gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp



On 3/17/2023 4:20 PM, Thinh Nguyen wrote:
> On Fri, Mar 17, 2023, Thinh Nguyen wrote:
>> On Fri, Mar 17, 2023, Elson Serrao wrote:
>>>
>>>
>>> On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
>>>> On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
>>>>> When host sends a suspend notification to the device, handle
>>>>> the suspend callbacks in the function driver. Enhanced super
>>>>> speed devices can support function suspend feature to put the
>>>>> function in suspend state. Handle function suspend callback.
>>>>>
>>>>> Depending on the remote wakeup capability the device can either
>>>>> trigger a remote wakeup or wait for the host initiated resume to
>>>>> start data transfer again.
>>>>>
>>>>> Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
>>>>> ---
>>>>>    drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
>>>>>    drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
>>>>>    drivers/usb/gadget/function/u_ether.h |  4 ++
>>>>>    3 files changed, 138 insertions(+)
>>>>>
>>>>> diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
>>>>> index a7ab30e..c43cd557 100644
>>>>> --- a/drivers/usb/gadget/function/f_ecm.c
>>>>> +++ b/drivers/usb/gadget/function/f_ecm.c
>>>>> @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
>>>>>    	usb_ep_disable(ecm->notify);
>>>>>    	ecm->notify->desc = NULL;
>>>>> +	f->func_suspended = false;
>>>>> +	f->func_wakeup_armed = false;
>>>>>    }
>>>>>    /*-------------------------------------------------------------------------*/
>>>>> @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
>>>>>    	return &opts->func_inst;
>>>>>    }
>>>>> +static void ecm_suspend(struct usb_function *f)
>>>>> +{
>>>>> +	struct f_ecm *ecm = func_to_ecm(f);
>>>>> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>>>>> +
>>>>> +	if (f->func_suspended) {
>>>>> +		DBG(cdev, "Function already suspended\n");
>>>>> +		return;
>>>>> +	}
>>>>> +
>>>>> +	DBG(cdev, "ECM Suspend\n");
>>>>> +
>>>>> +	gether_suspend(&ecm->port);
>>>>> +}
>>>>> +
>>>>> +static void ecm_resume(struct usb_function *f)
>>>>> +{
>>>>> +	struct f_ecm *ecm = func_to_ecm(f);
>>>>> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>>>>> +
>>>>> +	/*
>>>>> +	 * If the function is in USB3 Function Suspend state, resume is
>>>>> +	 * canceled. In this case resume is done by a Function Resume request.
>>>>> +	 */
>>>>> +	if (f->func_suspended)
>>>>> +		return;
>>>>> +
>>>>> +	DBG(cdev, "ECM Resume\n");
>>>>> +
>>>>> +	gether_resume(&ecm->port);
>>>>> +}
>>>>> +
>>>>> +static int ecm_get_status(struct usb_function *f)
>>>>> +{
>>>>> +	struct usb_configuration *c = f->config;
>>>>> +
>>>>> +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
>>>>> +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
>>>>> +		return 0;
>>>>> +
>>>>> +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
>>>>> +		USB_INTRF_STAT_FUNC_RW_CAP;
>>>>> +}
>>>>
>>>> Why do we need to implement ecm_get_status if it's already handled in
>>>> composite.c now?
>>>>
>>>
>>> Yes this can be removed now. Will modify accordingly.
>>>>> +
>>>>> +static int ecm_func_suspend(struct usb_function *f, u8 options)
>>>>> +{
>>>>> +	struct usb_composite_dev *cdev = f->config->cdev;
>>>>> +
>>>>> +	DBG(cdev, "func susp %u cmd\n", options);
>>>>> +
>>>>> +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
>>>>
>>>> This feature selector doesn't indicate whether it's SetFeature or
>>>> ClearFeature request. ecm_func_suspend is supposed to be for
>>>> SetFeature(suspend) only. Perhaps we may have to define func_resume()
>>>> for ClearFeature(suspend)?
>>>>
>>
>>> Host uses the same feature selector FUNCTION_SUSPEND for function suspend
>>> and function resume and func_suspend() callback can be used to
>>> handle both the cases ? The distinction comes whether it is a
>>
>> How do you plan to handle that? Pass this info in some unused/reserved
>> bit of the "options" argument? Introduce a new parameter to the
>> func_suspend()?
>>
>> If that's the case, then you need to update the document on
>> func_suspend() to also support ClearFeature(suspend). Right now it's
>> documented for SetFeature only. Also, make sure that other existing
>> function drivers will not break because of the change of the
>> func_suspend behavior.
>>
>>> SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) which can be
>>> easily done in the func_suspend callback itself. We can add another callback
>>> func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont that be
>>> redundant and more callback handling on function driver/composite side as
>>> well? Please let me know your opinion.
>>>
>>
>> We actually didn't properly define func_suspend and its counter part. It
>> seems cleaner to me to introduce func_resume as it seems more intuitive
>> and easier to read. Let me know how you plan to use func_suspend() for
>> both cases.
>>
> 
> How about we handle function suspend resume in composite also? I mean
> something like this:
> 
> diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
> index 36add1879ed2..79dc055eb5f7 100644
> --- a/drivers/usb/gadget/composite.c
> +++ b/drivers/usb/gadget/composite.c
> @@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
>   		f = cdev->config->interface[intf];
>   		if (!f)
>   			break;
> -		status = f->get_status ? f->get_status(f) : 0;
> -		if (status < 0)
> -			break;
> +
> +		if (f->get_status) {
> +			status = f->get_status(f);
> +			if (status < 0)
> +				break;
> +		} else {
> +			if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
> +				status |= USB_INTRF_STAT_FUNC_RW_CAP;
> +				if (f->func_wakeup_armed)
> +					status |= USB_INTRF_STAT_FUNC_RW;
> +			}
> +		}
>   		put_unaligned_le16(status & 0x0000ffff, req->buf);
>   		break;
>   	/*
> @@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
>   			f = cdev->config->interface[intf];
>   			if (!f)
>   				break;
> +			if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
> +				if (!(f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP))
> +					break;
> +
> +				f->func_wakeup_armed = (ctrl->bRequest == USB_REQ_SET_FEATURE);
> +			}
> +
>   			value = 0;
> -			if (f->func_suspend)
> +			if (f->func_suspend) {
>   				value = f->func_suspend(f, w_index >> 8);
> +			} else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
> +				if (f->suspend && && !f->func_suspended &&
> +				    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
> +					f->suspend(f);
> +					f->func_suspended = true;
> +				} else if (f->resume && f->func_suspended &&
> +					   ctrl->bRequest == USB_REQ_CLEAR_FEATURE_FEATURE)) {
> +					f->resume(f);
> +					f->func_suspended = false;
> +				}
> +			}
> +
>   			if (value < 0) {
>   				ERROR(cdev,
>   				      "func_suspend() returned error %d\n",
> 
At individual function driver level there is no need to differentiate 
between suspend() and func_suspend() APIs, as both are intended to put 
the function in suspend state. So your idea/implementation above makes 
it much more clearer. Let composite also handle this and call either 
f->suspend() or f->resume() callback based on the setup packet received. 
Thank you for this suggestion.

> 
> Also, do we need the f->func_suspended flag? we'd need the remote wakeup
> flag for the status, but when do we need f->func_suspended? It seems
> like it can be handled within the function driver's scope.

f->func_suspended flag I had added for below purposes

1.) Function drivers should know the right wakeup() op to be called.
That is if they are in FUNC_SUSPEND then call usb_func_wakeup() and if 
they are in device suspend then call usb_gadget_wakeup(). (we can use 
f->func_wakeup_armed flag for this purpose as well)

2.) If a function is in USB3 FUNCTION_SUSPEND state then it shouldn't 
allow f->resume() called through composite_resume() as the exit from 
FUNCTION_SUSPEND state is via ClearFeature(FUNCTION_SUSPEND).

So we need a way to tell function drivers if they are in USB3 
FUNCTION_SUSPEND state OR device suspend.

Please let me know if you see any alternative or better approach here.

Thanks
Elson

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-18  0:23           ` Elson Serrao
@ 2023-03-18  2:26             ` Thinh Nguyen
  2023-03-20 17:42               ` Elson Serrao
  0 siblings, 1 reply; 18+ messages in thread
From: Thinh Nguyen @ 2023-03-18  2:26 UTC (permalink / raw)
  To: Elson Serrao
  Cc: Thinh Nguyen, gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp

On Fri, Mar 17, 2023, Elson Serrao wrote:
> 
> 
> On 3/17/2023 4:20 PM, Thinh Nguyen wrote:
> > On Fri, Mar 17, 2023, Thinh Nguyen wrote:
> > > On Fri, Mar 17, 2023, Elson Serrao wrote:
> > > > 
> > > > 
> > > > On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
> > > > > On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
> > > > > > When host sends a suspend notification to the device, handle
> > > > > > the suspend callbacks in the function driver. Enhanced super
> > > > > > speed devices can support function suspend feature to put the
> > > > > > function in suspend state. Handle function suspend callback.
> > > > > > 
> > > > > > Depending on the remote wakeup capability the device can either
> > > > > > trigger a remote wakeup or wait for the host initiated resume to
> > > > > > start data transfer again.
> > > > > > 
> > > > > > Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
> > > > > > ---
> > > > > >    drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
> > > > > >    drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
> > > > > >    drivers/usb/gadget/function/u_ether.h |  4 ++
> > > > > >    3 files changed, 138 insertions(+)
> > > > > > 
> > > > > > diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
> > > > > > index a7ab30e..c43cd557 100644
> > > > > > --- a/drivers/usb/gadget/function/f_ecm.c
> > > > > > +++ b/drivers/usb/gadget/function/f_ecm.c
> > > > > > @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
> > > > > >    	usb_ep_disable(ecm->notify);
> > > > > >    	ecm->notify->desc = NULL;
> > > > > > +	f->func_suspended = false;
> > > > > > +	f->func_wakeup_armed = false;
> > > > > >    }
> > > > > >    /*-------------------------------------------------------------------------*/
> > > > > > @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
> > > > > >    	return &opts->func_inst;
> > > > > >    }
> > > > > > +static void ecm_suspend(struct usb_function *f)
> > > > > > +{
> > > > > > +	struct f_ecm *ecm = func_to_ecm(f);
> > > > > > +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > > > +
> > > > > > +	if (f->func_suspended) {
> > > > > > +		DBG(cdev, "Function already suspended\n");
> > > > > > +		return;
> > > > > > +	}
> > > > > > +
> > > > > > +	DBG(cdev, "ECM Suspend\n");
> > > > > > +
> > > > > > +	gether_suspend(&ecm->port);
> > > > > > +}
> > > > > > +
> > > > > > +static void ecm_resume(struct usb_function *f)
> > > > > > +{
> > > > > > +	struct f_ecm *ecm = func_to_ecm(f);
> > > > > > +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > > > +
> > > > > > +	/*
> > > > > > +	 * If the function is in USB3 Function Suspend state, resume is
> > > > > > +	 * canceled. In this case resume is done by a Function Resume request.
> > > > > > +	 */
> > > > > > +	if (f->func_suspended)
> > > > > > +		return;
> > > > > > +
> > > > > > +	DBG(cdev, "ECM Resume\n");
> > > > > > +
> > > > > > +	gether_resume(&ecm->port);
> > > > > > +}
> > > > > > +
> > > > > > +static int ecm_get_status(struct usb_function *f)
> > > > > > +{
> > > > > > +	struct usb_configuration *c = f->config;
> > > > > > +
> > > > > > +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
> > > > > > +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
> > > > > > +		return 0;
> > > > > > +
> > > > > > +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
> > > > > > +		USB_INTRF_STAT_FUNC_RW_CAP;
> > > > > > +}
> > > > > 
> > > > > Why do we need to implement ecm_get_status if it's already handled in
> > > > > composite.c now?
> > > > > 
> > > > 
> > > > Yes this can be removed now. Will modify accordingly.
> > > > > > +
> > > > > > +static int ecm_func_suspend(struct usb_function *f, u8 options)
> > > > > > +{
> > > > > > +	struct usb_composite_dev *cdev = f->config->cdev;
> > > > > > +
> > > > > > +	DBG(cdev, "func susp %u cmd\n", options);
> > > > > > +
> > > > > > +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
> > > > > 
> > > > > This feature selector doesn't indicate whether it's SetFeature or
> > > > > ClearFeature request. ecm_func_suspend is supposed to be for
> > > > > SetFeature(suspend) only. Perhaps we may have to define func_resume()
> > > > > for ClearFeature(suspend)?
> > > > > 
> > > 
> > > > Host uses the same feature selector FUNCTION_SUSPEND for function suspend
> > > > and function resume and func_suspend() callback can be used to
> > > > handle both the cases ? The distinction comes whether it is a
> > > 
> > > How do you plan to handle that? Pass this info in some unused/reserved
> > > bit of the "options" argument? Introduce a new parameter to the
> > > func_suspend()?
> > > 
> > > If that's the case, then you need to update the document on
> > > func_suspend() to also support ClearFeature(suspend). Right now it's
> > > documented for SetFeature only. Also, make sure that other existing
> > > function drivers will not break because of the change of the
> > > func_suspend behavior.
> > > 
> > > > SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) which can be
> > > > easily done in the func_suspend callback itself. We can add another callback
> > > > func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont that be
> > > > redundant and more callback handling on function driver/composite side as
> > > > well? Please let me know your opinion.
> > > > 
> > > 
> > > We actually didn't properly define func_suspend and its counter part. It
> > > seems cleaner to me to introduce func_resume as it seems more intuitive
> > > and easier to read. Let me know how you plan to use func_suspend() for
> > > both cases.
> > > 
> > 
> > How about we handle function suspend resume in composite also? I mean
> > something like this:
> > 
> > diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
> > index 36add1879ed2..79dc055eb5f7 100644
> > --- a/drivers/usb/gadget/composite.c
> > +++ b/drivers/usb/gadget/composite.c
> > @@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
> >   		f = cdev->config->interface[intf];
> >   		if (!f)
> >   			break;
> > -		status = f->get_status ? f->get_status(f) : 0;
> > -		if (status < 0)
> > -			break;
> > +
> > +		if (f->get_status) {
> > +			status = f->get_status(f);
> > +			if (status < 0)
> > +				break;
> > +		} else {
> > +			if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
> > +				status |= USB_INTRF_STAT_FUNC_RW_CAP;
> > +				if (f->func_wakeup_armed)
> > +					status |= USB_INTRF_STAT_FUNC_RW;
> > +			}
> > +		}
> >   		put_unaligned_le16(status & 0x0000ffff, req->buf);
> >   		break;
> >   	/*
> > @@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
> >   			f = cdev->config->interface[intf];
> >   			if (!f)
> >   				break;
> > +			if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
> > +				if (!(f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP))
> > +					break;
> > +
> > +				f->func_wakeup_armed = (ctrl->bRequest == USB_REQ_SET_FEATURE);
> > +			}
> > +
> >   			value = 0;
> > -			if (f->func_suspend)
> > +			if (f->func_suspend) {
> >   				value = f->func_suspend(f, w_index >> 8);
> > +			} else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
> > +				if (f->suspend && && !f->func_suspended &&
> > +				    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
> > +					f->suspend(f);
> > +					f->func_suspended = true;
> > +				} else if (f->resume && f->func_suspended &&
> > +					   ctrl->bRequest == USB_REQ_CLEAR_FEATURE_FEATURE)) {
> > +					f->resume(f);
> > +					f->func_suspended = false;
> > +				}
> > +			}
> > +
> >   			if (value < 0) {
> >   				ERROR(cdev,
> >   				      "func_suspend() returned error %d\n",
> > 
> At individual function driver level there is no need to differentiate
> between suspend() and func_suspend() APIs, as both are intended to put the
> function in suspend state. So your idea/implementation above makes it much
> more clearer. Let composite also handle this and call either f->suspend() or
> f->resume() callback based on the setup packet received. Thank you for this
> suggestion.
> 
> > 
> > Also, do we need the f->func_suspended flag? we'd need the remote wakeup
> > flag for the status, but when do we need f->func_suspended? It seems
> > like it can be handled within the function driver's scope.
> 
> f->func_suspended flag I had added for below purposes
> 
> 1.) Function drivers should know the right wakeup() op to be called.
> That is if they are in FUNC_SUSPEND then call usb_func_wakeup() and if they
> are in device suspend then call usb_gadget_wakeup(). (we can use
> f->func_wakeup_armed flag for this purpose as well)
> 
> 2.) If a function is in USB3 FUNCTION_SUSPEND state then it shouldn't allow
> f->resume() called through composite_resume() as the exit from
> FUNCTION_SUSPEND state is via ClearFeature(FUNCTION_SUSPEND).
> 
> So we need a way to tell function drivers if they are in USB3
> FUNCTION_SUSPEND state OR device suspend.

Ok. So do you plan to update composite_resume to check that? Perhaps
document this expected behavior also?

> 
> Please let me know if you see any alternative or better approach here.
> 

If we have a use for it in composite.c, then it should be made
available. Otherwise it should only exist in the function driver (ie. in
f_ecm structure).

Thanks,
Thinh

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-18  2:26             ` Thinh Nguyen
@ 2023-03-20 17:42               ` Elson Serrao
  2023-03-24  0:12                 ` Elson Serrao
  0 siblings, 1 reply; 18+ messages in thread
From: Elson Serrao @ 2023-03-20 17:42 UTC (permalink / raw)
  To: Thinh Nguyen; +Cc: gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp



On 3/17/2023 7:26 PM, Thinh Nguyen wrote:
> On Fri, Mar 17, 2023, Elson Serrao wrote:
>>
>>
>> On 3/17/2023 4:20 PM, Thinh Nguyen wrote:
>>> On Fri, Mar 17, 2023, Thinh Nguyen wrote:
>>>> On Fri, Mar 17, 2023, Elson Serrao wrote:
>>>>>
>>>>>
>>>>> On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
>>>>>> On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
>>>>>>> When host sends a suspend notification to the device, handle
>>>>>>> the suspend callbacks in the function driver. Enhanced super
>>>>>>> speed devices can support function suspend feature to put the
>>>>>>> function in suspend state. Handle function suspend callback.
>>>>>>>
>>>>>>> Depending on the remote wakeup capability the device can either
>>>>>>> trigger a remote wakeup or wait for the host initiated resume to
>>>>>>> start data transfer again.
>>>>>>>
>>>>>>> Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
>>>>>>> ---
>>>>>>>     drivers/usb/gadget/function/f_ecm.c   | 71 +++++++++++++++++++++++++++++++++++
>>>>>>>     drivers/usb/gadget/function/u_ether.c | 63 +++++++++++++++++++++++++++++++
>>>>>>>     drivers/usb/gadget/function/u_ether.h |  4 ++
>>>>>>>     3 files changed, 138 insertions(+)
>>>>>>>
>>>>>>> diff --git a/drivers/usb/gadget/function/f_ecm.c b/drivers/usb/gadget/function/f_ecm.c
>>>>>>> index a7ab30e..c43cd557 100644
>>>>>>> --- a/drivers/usb/gadget/function/f_ecm.c
>>>>>>> +++ b/drivers/usb/gadget/function/f_ecm.c
>>>>>>> @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
>>>>>>>     	usb_ep_disable(ecm->notify);
>>>>>>>     	ecm->notify->desc = NULL;
>>>>>>> +	f->func_suspended = false;
>>>>>>> +	f->func_wakeup_armed = false;
>>>>>>>     }
>>>>>>>     /*-------------------------------------------------------------------------*/
>>>>>>> @@ -885,6 +887,71 @@ static struct usb_function_instance *ecm_alloc_inst(void)
>>>>>>>     	return &opts->func_inst;
>>>>>>>     }
>>>>>>> +static void ecm_suspend(struct usb_function *f)
>>>>>>> +{
>>>>>>> +	struct f_ecm *ecm = func_to_ecm(f);
>>>>>>> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>>>>>>> +
>>>>>>> +	if (f->func_suspended) {
>>>>>>> +		DBG(cdev, "Function already suspended\n");
>>>>>>> +		return;
>>>>>>> +	}
>>>>>>> +
>>>>>>> +	DBG(cdev, "ECM Suspend\n");
>>>>>>> +
>>>>>>> +	gether_suspend(&ecm->port);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static void ecm_resume(struct usb_function *f)
>>>>>>> +{
>>>>>>> +	struct f_ecm *ecm = func_to_ecm(f);
>>>>>>> +	struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>>>>>>> +
>>>>>>> +	/*
>>>>>>> +	 * If the function is in USB3 Function Suspend state, resume is
>>>>>>> +	 * canceled. In this case resume is done by a Function Resume request.
>>>>>>> +	 */
>>>>>>> +	if (f->func_suspended)
>>>>>>> +		return;
>>>>>>> +
>>>>>>> +	DBG(cdev, "ECM Resume\n");
>>>>>>> +
>>>>>>> +	gether_resume(&ecm->port);
>>>>>>> +}
>>>>>>> +
>>>>>>> +static int ecm_get_status(struct usb_function *f)
>>>>>>> +{
>>>>>>> +	struct usb_configuration *c = f->config;
>>>>>>> +
>>>>>>> +	/* D0 and D1 bit set to 0 if device is not wakeup capable */
>>>>>>> +	if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
>>>>>>> +		return 0;
>>>>>>> +
>>>>>>> +	return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
>>>>>>> +		USB_INTRF_STAT_FUNC_RW_CAP;
>>>>>>> +}
>>>>>>
>>>>>> Why do we need to implement ecm_get_status if it's already handled in
>>>>>> composite.c now?
>>>>>>
>>>>>
>>>>> Yes this can be removed now. Will modify accordingly.
>>>>>>> +
>>>>>>> +static int ecm_func_suspend(struct usb_function *f, u8 options)
>>>>>>> +{
>>>>>>> +	struct usb_composite_dev *cdev = f->config->cdev;
>>>>>>> +
>>>>>>> +	DBG(cdev, "func susp %u cmd\n", options);
>>>>>>> +
>>>>>>> +	if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
>>>>>>
>>>>>> This feature selector doesn't indicate whether it's SetFeature or
>>>>>> ClearFeature request. ecm_func_suspend is supposed to be for
>>>>>> SetFeature(suspend) only. Perhaps we may have to define func_resume()
>>>>>> for ClearFeature(suspend)?
>>>>>>
>>>>
>>>>> Host uses the same feature selector FUNCTION_SUSPEND for function suspend
>>>>> and function resume and func_suspend() callback can be used to
>>>>> handle both the cases ? The distinction comes whether it is a
>>>>
>>>> How do you plan to handle that? Pass this info in some unused/reserved
>>>> bit of the "options" argument? Introduce a new parameter to the
>>>> func_suspend()?
>>>>
>>>> If that's the case, then you need to update the document on
>>>> func_suspend() to also support ClearFeature(suspend). Right now it's
>>>> documented for SetFeature only. Also, make sure that other existing
>>>> function drivers will not break because of the change of the
>>>> func_suspend behavior.
>>>>
>>>>> SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) which can be
>>>>> easily done in the func_suspend callback itself. We can add another callback
>>>>> func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont that be
>>>>> redundant and more callback handling on function driver/composite side as
>>>>> well? Please let me know your opinion.
>>>>>
>>>>
>>>> We actually didn't properly define func_suspend and its counter part. It
>>>> seems cleaner to me to introduce func_resume as it seems more intuitive
>>>> and easier to read. Let me know how you plan to use func_suspend() for
>>>> both cases.
>>>>
>>>
>>> How about we handle function suspend resume in composite also? I mean
>>> something like this:
>>>
>>> diff --git a/drivers/usb/gadget/composite.c b/drivers/usb/gadget/composite.c
>>> index 36add1879ed2..79dc055eb5f7 100644
>>> --- a/drivers/usb/gadget/composite.c
>>> +++ b/drivers/usb/gadget/composite.c
>>> @@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
>>>    		f = cdev->config->interface[intf];
>>>    		if (!f)
>>>    			break;
>>> -		status = f->get_status ? f->get_status(f) : 0;
>>> -		if (status < 0)
>>> -			break;
>>> +
>>> +		if (f->get_status) {
>>> +			status = f->get_status(f);
>>> +			if (status < 0)
>>> +				break;
>>> +		} else {
>>> +			if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
>>> +				status |= USB_INTRF_STAT_FUNC_RW_CAP;
>>> +				if (f->func_wakeup_armed)
>>> +					status |= USB_INTRF_STAT_FUNC_RW;
>>> +			}
>>> +		}
>>>    		put_unaligned_le16(status & 0x0000ffff, req->buf);
>>>    		break;
>>>    	/*
>>> @@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
>>>    			f = cdev->config->interface[intf];
>>>    			if (!f)
>>>    				break;
>>> +			if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
>>> +				if (!(f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP))
>>> +					break;
>>> +
>>> +				f->func_wakeup_armed = (ctrl->bRequest == USB_REQ_SET_FEATURE);
>>> +			}
>>> +
>>>    			value = 0;
>>> -			if (f->func_suspend)
>>> +			if (f->func_suspend) {
>>>    				value = f->func_suspend(f, w_index >> 8);
>>> +			} else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
>>> +				if (f->suspend && && !f->func_suspended &&
>>> +				    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
>>> +					f->suspend(f);
>>> +					f->func_suspended = true;
>>> +				} else if (f->resume && f->func_suspended &&
>>> +					   ctrl->bRequest == USB_REQ_CLEAR_FEATURE_FEATURE)) {
>>> +					f->resume(f);
>>> +					f->func_suspended = false;
>>> +				}
>>> +			}
>>> +
>>>    			if (value < 0) {
>>>    				ERROR(cdev,
>>>    				      "func_suspend() returned error %d\n",
>>>
>> At individual function driver level there is no need to differentiate
>> between suspend() and func_suspend() APIs, as both are intended to put the
>> function in suspend state. So your idea/implementation above makes it much
>> more clearer. Let composite also handle this and call either f->suspend() or
>> f->resume() callback based on the setup packet received. Thank you for this
>> suggestion.
>>
>>>
>>> Also, do we need the f->func_suspended flag? we'd need the remote wakeup
>>> flag for the status, but when do we need f->func_suspended? It seems
>>> like it can be handled within the function driver's scope.
>>
>> f->func_suspended flag I had added for below purposes
>>
>> 1.) Function drivers should know the right wakeup() op to be called.
>> That is if they are in FUNC_SUSPEND then call usb_func_wakeup() and if they
>> are in device suspend then call usb_gadget_wakeup(). (we can use
>> f->func_wakeup_armed flag for this purpose as well)
>>
>> 2.) If a function is in USB3 FUNCTION_SUSPEND state then it shouldn't allow
>> f->resume() called through composite_resume() as the exit from
>> FUNCTION_SUSPEND state is via ClearFeature(FUNCTION_SUSPEND).
>>
>> So we need a way to tell function drivers if they are in USB3
>> FUNCTION_SUSPEND state OR device suspend.
> 
> Ok. So do you plan to update composite_resume to check that? Perhaps
> document this expected behavior also?
> 
Yes I will update composite resume to handle this scenario. I was 
handling it in f_ecm till now, but based on this new discussion I will 
move all handling to composite layer.
>>
>> Please let me know if you see any alternative or better approach here.
>>
> 
> If we have a use for it in composite.c, then it should be made
> available. Otherwise it should only exist in the function driver (ie. in
> f_ecm structure).
> 
Yes I will keep it for now. Since we are moving all handling to 
composite layer better to have some kind of flag that function 
drivers/composite layer can use to check if a given function is in 
FUNCTION_SUSPEND state or composite_suspend() state.

Thanks
Elson


^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-16 23:38 ` [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support Elson Roy Serrao
  2023-03-17  0:11   ` Thinh Nguyen
@ 2023-03-21  8:12   ` Dan Carpenter
  1 sibling, 0 replies; 18+ messages in thread
From: Dan Carpenter @ 2023-03-21  8:12 UTC (permalink / raw)
  To: oe-kbuild, Elson Roy Serrao, gregkh, Thinh.Nguyen
  Cc: lkp, oe-kbuild-all, linux-kernel, linux-usb, quic_wcheng,
	quic_jackp, Elson Roy Serrao

Hi Elson,

https://git-scm.com/docs/git-format-patch#_base_tree_information]

url:    https://github.com/intel-lab-lkp/linux/commits/Elson-Roy-Serrao/usb-gadget-Properly-configure-the-device-for-remote-wakeup/20230317-074030
base:   https://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb.git usb-testing
patch link:    https://lore.kernel.org/r/1679009888-8239-7-git-send-email-quic_eserrao%40quicinc.com
patch subject: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
config: riscv-randconfig-m031-20230319 (https://download.01.org/0day-ci/archive/20230321/202303211515.XaO8YKCz-lkp@intel.com/config)
compiler: riscv32-linux-gcc (GCC) 12.1.0

If you fix the issue, kindly add following tag where applicable
| Reported-by: kernel test robot <lkp@intel.com>
| Reported-by: Dan Carpenter <error27@gmail.com>
| Link: https://lore.kernel.org/r/202303211515.XaO8YKCz-lkp@intel.com/

New smatch warnings:
drivers/usb/gadget/function/u_ether.c:474 eth_start_xmit() error: we previously assumed 'dev->port_usb' could be null (see line 466)
drivers/usb/gadget/function/u_ether.c:539 eth_start_xmit() warn: variable dereferenced before check 'dev->port_usb' (see line 474)

Old smatch warnings:
drivers/usb/gadget/function/u_ether.c:553 eth_start_xmit() error: we previously assumed 'skb' could be null (see line 491)

vim +474 drivers/usb/gadget/function/u_ether.c

25a79c41ce0ce8 drivers/usb/gadget/u_ether.c          Stephen Hemminger   2009-08-31  454  static netdev_tx_t eth_start_xmit(struct sk_buff *skb,
25a79c41ce0ce8 drivers/usb/gadget/u_ether.c          Stephen Hemminger   2009-08-31  455  					struct net_device *net)
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  456  {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  457  	struct eth_dev		*dev = netdev_priv(net);
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  458  	int			length = 0;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  459  	int			retval;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  460  	struct usb_request	*req = NULL;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  461  	unsigned long		flags;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  462  	struct usb_ep		*in;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  463  	u16			cdc_filter;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  464  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  465  	spin_lock_irqsave(&dev->lock, flags);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19 @466  	if (dev->port_usb) {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  467  		in = dev->port_usb->in_ep;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  468  		cdc_filter = dev->port_usb->cdc_filter;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  469  	} else {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  470  		in = NULL;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  471  		cdc_filter = 0;

NULL on this path.

2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  472  	}
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  473  
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16 @474  	if (dev->port_usb->is_suspend) {
                                                                                                    ^^^^^^^^^^^^^^^
Dead.  (Both ->port_usb warnings are caused by this dereference).

11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  475  		DBG(dev, "Port suspended. Triggering wakeup\n");
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  476  		netif_stop_queue(net);
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  477  		spin_unlock_irqrestore(&dev->lock, flags);
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  478  		ether_wakeup_host(dev->port_usb);
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  479  		return NETDEV_TX_BUSY;
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  480  	}
11f254fc796c87 drivers/usb/gadget/function/u_ether.c Elson Roy Serrao    2023-03-16  481  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  482  	spin_unlock_irqrestore(&dev->lock, flags);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  483  
8ae01239609b29 drivers/usb/gadget/function/u_ether.c Maciej Żenczykowski 2021-07-01  484  	if (!in) {
8ae01239609b29 drivers/usb/gadget/function/u_ether.c Maciej Żenczykowski 2021-07-01  485  		if (skb)
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  486  			dev_kfree_skb_any(skb);
6ed106549d1747 drivers/usb/gadget/u_ether.c          Patrick McHardy     2009-06-23  487  		return NETDEV_TX_OK;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  488  	}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  489  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  490  	/* apply outgoing CDC or RNDIS filters */
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  491  	if (skb && !is_promisc(cdc_filter)) {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  492  		u8		*dest = skb->data;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  493  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  494  		if (is_multicast_ether_addr(dest)) {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  495  			u16	type;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  496  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  497  			/* ignores USB_CDC_PACKET_TYPE_MULTICAST and host
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  498  			 * SET_ETHERNET_MULTICAST_FILTERS requests
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  499  			 */
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  500  			if (is_broadcast_ether_addr(dest))
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  501  				type = USB_CDC_PACKET_TYPE_BROADCAST;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  502  			else
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  503  				type = USB_CDC_PACKET_TYPE_ALL_MULTICAST;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  504  			if (!(cdc_filter & type)) {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  505  				dev_kfree_skb_any(skb);
6ed106549d1747 drivers/usb/gadget/u_ether.c          Patrick McHardy     2009-06-23  506  				return NETDEV_TX_OK;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  507  			}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  508  		}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  509  		/* ignores USB_CDC_PACKET_TYPE_DIRECTED */
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  510  	}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  511  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  512  	spin_lock_irqsave(&dev->req_lock, flags);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  513  	/*
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  514  	 * this freelist can be empty if an interrupt triggered disconnect()
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  515  	 * and reconfigured the gadget (shutting down this queue) after the
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  516  	 * network stack decided to xmit but before we got the spinlock.
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  517  	 */
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  518  	if (list_empty(&dev->tx_reqs)) {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  519  		spin_unlock_irqrestore(&dev->req_lock, flags);
5b548140225c6b drivers/usb/gadget/u_ether.c          Patrick McHardy     2009-06-12  520  		return NETDEV_TX_BUSY;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  521  	}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  522  
fea14e68ff5e11 drivers/usb/gadget/function/u_ether.c Felipe Balbi        2017-03-22  523  	req = list_first_entry(&dev->tx_reqs, struct usb_request, list);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  524  	list_del(&req->list);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  525  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  526  	/* temporarily stop TX queue when the freelist empties */
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  527  	if (list_empty(&dev->tx_reqs))
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  528  		netif_stop_queue(net);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  529  	spin_unlock_irqrestore(&dev->req_lock, flags);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  530  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  531  	/* no buffer copies needed, unless the network stack did it
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  532  	 * or the hardware can't use skb buffers.
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  533  	 * or there's not enough space for extra headers we need
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  534  	 */
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  535  	if (dev->wrap) {
9b39e9ddedeef4 drivers/usb/gadget/u_ether.c          Brian Niebuhr       2009-08-14  536  		unsigned long	flags;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  537  
9b39e9ddedeef4 drivers/usb/gadget/u_ether.c          Brian Niebuhr       2009-08-14  538  		spin_lock_irqsave(&dev->lock, flags);
9b39e9ddedeef4 drivers/usb/gadget/u_ether.c          Brian Niebuhr       2009-08-14 @539  		if (dev->port_usb)
9b39e9ddedeef4 drivers/usb/gadget/u_ether.c          Brian Niebuhr       2009-08-14  540  			skb = dev->wrap(dev->port_usb, skb);
3a383cc0b8cc33 drivers/usb/gadget/function/u_ether.c Greg Kroah-Hartman  2016-09-19  541  		spin_unlock_irqrestore(&dev->lock, flags);
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  542  		if (!skb) {
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  543  			/* Multi frame CDC protocols may store the frame for
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  544  			 * later which is not a dropped frame.
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  545  			 */
88c09eacf560c3 drivers/usb/gadget/function/u_ether.c Peter Chen          2016-07-01  546  			if (dev->port_usb &&
3a383cc0b8cc33 drivers/usb/gadget/function/u_ether.c Greg Kroah-Hartman  2016-09-19  547  					dev->port_usb->supports_multi_frame)
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  548  				goto multiframe;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  549  			goto drop;
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  550  		}
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  551  	}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  552  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  553  	length = skb->len;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  554  	req->buf = skb->data;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  555  	req->context = skb;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  556  	req->complete = tx_complete;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  557  
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  558  	/* NCM requires no zlp if transfer is dwNtbInMaxSize */
79775f44183840 drivers/usb/gadget/function/u_ether.c Harish Jenny K N    2016-09-09  559  	if (dev->port_usb &&
79775f44183840 drivers/usb/gadget/function/u_ether.c Harish Jenny K N    2016-09-09  560  	    dev->port_usb->is_fixed &&
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  561  	    length == dev->port_usb->fixed_in_len &&
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  562  	    (length % in->maxpacket) == 0)
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  563  		req->zero = 0;
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  564  	else
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  565  		req->zero = 1;
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  566  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  567  	/* use zlp framing on tx for strict CDC-Ether conformance,
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  568  	 * though any robust network rx path ignores extra padding.
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  569  	 * and some hardware doesn't like to write zlps.
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  570  	 */
5c1168dbc50828 drivers/usb/gadget/u_ether.c          Yauheni Kaliuta     2010-12-08  571  	if (req->zero && !dev->zlp && (length % in->maxpacket) == 0)
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  572  		length++;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  573  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  574  	req->length = length;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  575  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  576  	retval = usb_ep_queue(in, req, GFP_ATOMIC);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  577  	switch (retval) {
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  578  	default:
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  579  		DBG(dev, "tx queue err %d\n", retval);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  580  		break;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  581  	case 0:
860e9538a9482b drivers/usb/gadget/function/u_ether.c Florian Westphal    2016-05-03  582  		netif_trans_update(net);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  583  		atomic_inc(&dev->tx_qlen);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  584  	}
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  585  
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  586  	if (retval) {
9b39e9ddedeef4 drivers/usb/gadget/u_ether.c          Brian Niebuhr       2009-08-14  587  		dev_kfree_skb_any(skb);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  588  drop:
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  589  		dev->net->stats.tx_dropped++;
6d3865f9d41f15 drivers/usb/gadget/u_ether.c          Jim Baxter          2014-07-07  590  multiframe:
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  591  		spin_lock_irqsave(&dev->req_lock, flags);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  592  		if (list_empty(&dev->tx_reqs))
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  593  			netif_start_queue(net);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  594  		list_add(&req->list, &dev->tx_reqs);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  595  		spin_unlock_irqrestore(&dev->req_lock, flags);
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  596  	}
6ed106549d1747 drivers/usb/gadget/u_ether.c          Patrick McHardy     2009-06-23  597  	return NETDEV_TX_OK;
2b3d942c487808 drivers/usb/gadget/u_ether.c          David Brownell      2008-06-19  598  }

-- 
0-DAY CI Kernel Test Service
https://github.com/intel/lkp-tests


^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-20 17:42               ` Elson Serrao
@ 2023-03-24  0:12                 ` Elson Serrao
  2023-03-24  0:59                   ` Thinh Nguyen
  0 siblings, 1 reply; 18+ messages in thread
From: Elson Serrao @ 2023-03-24  0:12 UTC (permalink / raw)
  To: Thinh Nguyen; +Cc: gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp



On 3/20/2023 10:42 AM, Elson Serrao wrote:
> 
> 
> On 3/17/2023 7:26 PM, Thinh Nguyen wrote:
>> On Fri, Mar 17, 2023, Elson Serrao wrote:
>>>
>>>
>>> On 3/17/2023 4:20 PM, Thinh Nguyen wrote:
>>>> On Fri, Mar 17, 2023, Thinh Nguyen wrote:
>>>>> On Fri, Mar 17, 2023, Elson Serrao wrote:
>>>>>>
>>>>>>
>>>>>> On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
>>>>>>> On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
>>>>>>>> When host sends a suspend notification to the device, handle
>>>>>>>> the suspend callbacks in the function driver. Enhanced super
>>>>>>>> speed devices can support function suspend feature to put the
>>>>>>>> function in suspend state. Handle function suspend callback.
>>>>>>>>
>>>>>>>> Depending on the remote wakeup capability the device can either
>>>>>>>> trigger a remote wakeup or wait for the host initiated resume to
>>>>>>>> start data transfer again.
>>>>>>>>
>>>>>>>> Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
>>>>>>>> ---
>>>>>>>>     drivers/usb/gadget/function/f_ecm.c   | 71 
>>>>>>>> +++++++++++++++++++++++++++++++++++
>>>>>>>>     drivers/usb/gadget/function/u_ether.c | 63 
>>>>>>>> +++++++++++++++++++++++++++++++
>>>>>>>>     drivers/usb/gadget/function/u_ether.h |  4 ++
>>>>>>>>     3 files changed, 138 insertions(+)
>>>>>>>>
>>>>>>>> diff --git a/drivers/usb/gadget/function/f_ecm.c 
>>>>>>>> b/drivers/usb/gadget/function/f_ecm.c
>>>>>>>> index a7ab30e..c43cd557 100644
>>>>>>>> --- a/drivers/usb/gadget/function/f_ecm.c
>>>>>>>> +++ b/drivers/usb/gadget/function/f_ecm.c
>>>>>>>> @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
>>>>>>>>         usb_ep_disable(ecm->notify);
>>>>>>>>         ecm->notify->desc = NULL;
>>>>>>>> +    f->func_suspended = false;
>>>>>>>> +    f->func_wakeup_armed = false;
>>>>>>>>     }
>>>>>>>>     
>>>>>>>> /*-------------------------------------------------------------------------*/ 
>>>>>>>>
>>>>>>>> @@ -885,6 +887,71 @@ static struct usb_function_instance 
>>>>>>>> *ecm_alloc_inst(void)
>>>>>>>>         return &opts->func_inst;
>>>>>>>>     }
>>>>>>>> +static void ecm_suspend(struct usb_function *f)
>>>>>>>> +{
>>>>>>>> +    struct f_ecm *ecm = func_to_ecm(f);
>>>>>>>> +    struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>>>>>>>> +
>>>>>>>> +    if (f->func_suspended) {
>>>>>>>> +        DBG(cdev, "Function already suspended\n");
>>>>>>>> +        return;
>>>>>>>> +    }
>>>>>>>> +
>>>>>>>> +    DBG(cdev, "ECM Suspend\n");
>>>>>>>> +
>>>>>>>> +    gether_suspend(&ecm->port);
>>>>>>>> +}
>>>>>>>> +
>>>>>>>> +static void ecm_resume(struct usb_function *f)
>>>>>>>> +{
>>>>>>>> +    struct f_ecm *ecm = func_to_ecm(f);
>>>>>>>> +    struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
>>>>>>>> +
>>>>>>>> +    /*
>>>>>>>> +     * If the function is in USB3 Function Suspend state, 
>>>>>>>> resume is
>>>>>>>> +     * canceled. In this case resume is done by a Function 
>>>>>>>> Resume request.
>>>>>>>> +     */
>>>>>>>> +    if (f->func_suspended)
>>>>>>>> +        return;
>>>>>>>> +
>>>>>>>> +    DBG(cdev, "ECM Resume\n");
>>>>>>>> +
>>>>>>>> +    gether_resume(&ecm->port);
>>>>>>>> +}
>>>>>>>> +
>>>>>>>> +static int ecm_get_status(struct usb_function *f)
>>>>>>>> +{
>>>>>>>> +    struct usb_configuration *c = f->config;
>>>>>>>> +
>>>>>>>> +    /* D0 and D1 bit set to 0 if device is not wakeup capable */
>>>>>>>> +    if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
>>>>>>>> +        return 0;
>>>>>>>> +
>>>>>>>> +    return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
>>>>>>>> +        USB_INTRF_STAT_FUNC_RW_CAP;
>>>>>>>> +}
>>>>>>>
>>>>>>> Why do we need to implement ecm_get_status if it's already 
>>>>>>> handled in
>>>>>>> composite.c now?
>>>>>>>
>>>>>>
>>>>>> Yes this can be removed now. Will modify accordingly.
>>>>>>>> +
>>>>>>>> +static int ecm_func_suspend(struct usb_function *f, u8 options)
>>>>>>>> +{
>>>>>>>> +    struct usb_composite_dev *cdev = f->config->cdev;
>>>>>>>> +
>>>>>>>> +    DBG(cdev, "func susp %u cmd\n", options);
>>>>>>>> +
>>>>>>>> +    if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
>>>>>>>
>>>>>>> This feature selector doesn't indicate whether it's SetFeature or
>>>>>>> ClearFeature request. ecm_func_suspend is supposed to be for
>>>>>>> SetFeature(suspend) only. Perhaps we may have to define 
>>>>>>> func_resume()
>>>>>>> for ClearFeature(suspend)?
>>>>>>>
>>>>>
>>>>>> Host uses the same feature selector FUNCTION_SUSPEND for function 
>>>>>> suspend
>>>>>> and function resume and func_suspend() callback can be used to
>>>>>> handle both the cases ? The distinction comes whether it is a
>>>>>
>>>>> How do you plan to handle that? Pass this info in some unused/reserved
>>>>> bit of the "options" argument? Introduce a new parameter to the
>>>>> func_suspend()?
>>>>>
>>>>> If that's the case, then you need to update the document on
>>>>> func_suspend() to also support ClearFeature(suspend). Right now it's
>>>>> documented for SetFeature only. Also, make sure that other existing
>>>>> function drivers will not break because of the change of the
>>>>> func_suspend behavior.
>>>>>
>>>>>> SetFeature(FUNCTION_SUSPEND) or ClearFeature(FUNCTION_SUSPEND) 
>>>>>> which can be
>>>>>> easily done in the func_suspend callback itself. We can add 
>>>>>> another callback
>>>>>> func_resume specific to ClearFeature(FUNCTION_SUSPEND) but wont 
>>>>>> that be
>>>>>> redundant and more callback handling on function driver/composite 
>>>>>> side as
>>>>>> well? Please let me know your opinion.
>>>>>>
>>>>>
>>>>> We actually didn't properly define func_suspend and its counter 
>>>>> part. It
>>>>> seems cleaner to me to introduce func_resume as it seems more 
>>>>> intuitive
>>>>> and easier to read. Let me know how you plan to use func_suspend() for
>>>>> both cases.
>>>>>
>>>>
>>>> How about we handle function suspend resume in composite also? I mean
>>>> something like this:
>>>>
>>>> diff --git a/drivers/usb/gadget/composite.c 
>>>> b/drivers/usb/gadget/composite.c
>>>> index 36add1879ed2..79dc055eb5f7 100644
>>>> --- a/drivers/usb/gadget/composite.c
>>>> +++ b/drivers/usb/gadget/composite.c
>>>> @@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget *gadget, 
>>>> const struct usb_ctrlrequest *ctrl)
>>>>            f = cdev->config->interface[intf];
>>>>            if (!f)
>>>>                break;
>>>> -        status = f->get_status ? f->get_status(f) : 0;
>>>> -        if (status < 0)
>>>> -            break;
>>>> +
>>>> +        if (f->get_status) {
>>>> +            status = f->get_status(f);
>>>> +            if (status < 0)
>>>> +                break;
>>>> +        } else {
>>>> +            if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
>>>> +                status |= USB_INTRF_STAT_FUNC_RW_CAP;
>>>> +                if (f->func_wakeup_armed)
>>>> +                    status |= USB_INTRF_STAT_FUNC_RW;
>>>> +            }
>>>> +        }
>>>>            put_unaligned_le16(status & 0x0000ffff, req->buf);
>>>>            break;
>>>>        /*
>>>> @@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget *gadget, 
>>>> const struct usb_ctrlrequest *ctrl)
>>>>                f = cdev->config->interface[intf];
>>>>                if (!f)
>>>>                    break;
>>>> +            if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
>>>> +                if (!(f->config->bmAttributes & 
>>>> USB_CONFIG_ATT_WAKEUP))
>>>> +                    break;
>>>> +
>>>> +                f->func_wakeup_armed = (ctrl->bRequest == 
>>>> USB_REQ_SET_FEATURE);
>>>> +            }
>>>> +
>>>>                value = 0;
>>>> -            if (f->func_suspend)
>>>> +            if (f->func_suspend) {
>>>>                    value = f->func_suspend(f, w_index >> 8);
>>>> +            } else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
>>>> +                if (f->suspend && && !f->func_suspended &&
>>>> +                    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
>>>> +                    f->suspend(f);
>>>> +                    f->func_suspended = true;
>>>> +                } else if (f->resume && f->func_suspended &&
>>>> +                       ctrl->bRequest == 
>>>> USB_REQ_CLEAR_FEATURE_FEATURE)) {

Linux based hosts, dont use ClearFeature(FUNCTION_SUSPEND) for function 
resume. They use SetFeature(FUNCTION_SUSPEND) itself but with Bit(0) i.e 
USB_INTRF_FUNC_SUSPEND_LP reset. So we may not be able to distinguish 
based on SET_FEATURE and CLEAR_FEATURE packets for function suspend and 
function resume. Instead we can generalize (w_index & 
USB_INTRF_FUNC_SUSPEND_LP) for function suspend and all other cases for 
function resume.

if (f->func_suspend) {
		value = f->func_suspend(f, w_index >> 8);
} else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
		if (f->suspend && !f->func_suspended) {
				f->suspend(f);
				f->func_suspended = true;
		}
} else {
		if (f->resume && f->func_suspended) {
				f->resume(f);
				f->func_suspended = false;
		}
}


Thanks
Elson

>>>> +                    f->resume(f);
>>>> +                    f->func_suspended = false;
>>>> +                }
>>>> +            }
>>>> +
>>>>                if (value < 0) {
>>>>                    ERROR(cdev,
>>>>                          "func_suspend() returned error %d\n",
>>>>
>>> At individual function driver level there is no need to differentiate
>>> between suspend() and func_suspend() APIs, as both are intended to 
>>> put the
>>> function in suspend state. So your idea/implementation above makes it 
>>> much
>>> more clearer. Let composite also handle this and call either 
>>> f->suspend() or
>>> f->resume() callback based on the setup packet received. Thank you 
>>> for this
>>> suggestion.
>>>
>>>>
>>>> Also, do we need the f->func_suspended flag? we'd need the remote 
>>>> wakeup
>>>> flag for the status, but when do we need f->func_suspended? It seems
>>>> like it can be handled within the function driver's scope.
>>>
>>> f->func_suspended flag I had added for below purposes
>>>
>>> 1.) Function drivers should know the right wakeup() op to be called.
>>> That is if they are in FUNC_SUSPEND then call usb_func_wakeup() and 
>>> if they
>>> are in device suspend then call usb_gadget_wakeup(). (we can use
>>> f->func_wakeup_armed flag for this purpose as well)
>>>
>>> 2.) If a function is in USB3 FUNCTION_SUSPEND state then it shouldn't 
>>> allow
>>> f->resume() called through composite_resume() as the exit from
>>> FUNCTION_SUSPEND state is via ClearFeature(FUNCTION_SUSPEND).
>>>
>>> So we need a way to tell function drivers if they are in USB3
>>> FUNCTION_SUSPEND state OR device suspend.
>>
>> Ok. So do you plan to update composite_resume to check that? Perhaps
>> document this expected behavior also?
>>
> Yes I will update composite resume to handle this scenario. I was 
> handling it in f_ecm till now, but based on this new discussion I will 
> move all handling to composite layer.
>>>
>>> Please let me know if you see any alternative or better approach here.
>>>
>>
>> If we have a use for it in composite.c, then it should be made
>> available. Otherwise it should only exist in the function driver (ie. in
>> f_ecm structure).
>>
> Yes I will keep it for now. Since we are moving all handling to 
> composite layer better to have some kind of flag that function 
> drivers/composite layer can use to check if a given function is in 
> FUNCTION_SUSPEND state or composite_suspend() state.
> 
> Thanks
> Elson
> 

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-24  0:12                 ` Elson Serrao
@ 2023-03-24  0:59                   ` Thinh Nguyen
  2023-03-24  1:14                     ` Thinh Nguyen
  0 siblings, 1 reply; 18+ messages in thread
From: Thinh Nguyen @ 2023-03-24  0:59 UTC (permalink / raw)
  To: Elson Serrao
  Cc: Thinh Nguyen, gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp

On Thu, Mar 23, 2023, Elson Serrao wrote:
> 
> 
> On 3/20/2023 10:42 AM, Elson Serrao wrote:
> > 
> > 
> > On 3/17/2023 7:26 PM, Thinh Nguyen wrote:
> > > On Fri, Mar 17, 2023, Elson Serrao wrote:
> > > > 
> > > > 
> > > > On 3/17/2023 4:20 PM, Thinh Nguyen wrote:
> > > > > On Fri, Mar 17, 2023, Thinh Nguyen wrote:
> > > > > > On Fri, Mar 17, 2023, Elson Serrao wrote:
> > > > > > > 
> > > > > > > 
> > > > > > > On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
> > > > > > > > On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
> > > > > > > > > When host sends a suspend notification to the device, handle
> > > > > > > > > the suspend callbacks in the function driver. Enhanced super
> > > > > > > > > speed devices can support function suspend feature to put the
> > > > > > > > > function in suspend state. Handle function suspend callback.
> > > > > > > > > 
> > > > > > > > > Depending on the remote wakeup capability the device can either
> > > > > > > > > trigger a remote wakeup or wait for the host initiated resume to
> > > > > > > > > start data transfer again.
> > > > > > > > > 
> > > > > > > > > Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
> > > > > > > > > ---
> > > > > > > > >     drivers/usb/gadget/function/f_ecm.c   |
> > > > > > > > > 71 +++++++++++++++++++++++++++++++++++
> > > > > > > > >     drivers/usb/gadget/function/u_ether.c |
> > > > > > > > > 63 +++++++++++++++++++++++++++++++
> > > > > > > > >     drivers/usb/gadget/function/u_ether.h |  4 ++
> > > > > > > > >     3 files changed, 138 insertions(+)
> > > > > > > > > 
> > > > > > > > > diff --git
> > > > > > > > > a/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > b/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > index a7ab30e..c43cd557 100644
> > > > > > > > > --- a/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > +++ b/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
> > > > > > > > >         usb_ep_disable(ecm->notify);
> > > > > > > > >         ecm->notify->desc = NULL;
> > > > > > > > > +    f->func_suspended = false;
> > > > > > > > > +    f->func_wakeup_armed = false;
> > > > > > > > >     }
> > > > > > > > >      /*-------------------------------------------------------------------------*/
> > > > > > > > > 
> > > > > > > > > @@ -885,6 +887,71 @@ static struct
> > > > > > > > > usb_function_instance *ecm_alloc_inst(void)
> > > > > > > > >         return &opts->func_inst;
> > > > > > > > >     }
> > > > > > > > > +static void ecm_suspend(struct usb_function *f)
> > > > > > > > > +{
> > > > > > > > > +    struct f_ecm *ecm = func_to_ecm(f);
> > > > > > > > > +    struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > > > > > > +
> > > > > > > > > +    if (f->func_suspended) {
> > > > > > > > > +        DBG(cdev, "Function already suspended\n");
> > > > > > > > > +        return;
> > > > > > > > > +    }
> > > > > > > > > +
> > > > > > > > > +    DBG(cdev, "ECM Suspend\n");
> > > > > > > > > +
> > > > > > > > > +    gether_suspend(&ecm->port);
> > > > > > > > > +}
> > > > > > > > > +
> > > > > > > > > +static void ecm_resume(struct usb_function *f)
> > > > > > > > > +{
> > > > > > > > > +    struct f_ecm *ecm = func_to_ecm(f);
> > > > > > > > > +    struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > > > > > > +
> > > > > > > > > +    /*
> > > > > > > > > +     * If the function is in USB3 Function
> > > > > > > > > Suspend state, resume is
> > > > > > > > > +     * canceled. In this case resume is
> > > > > > > > > done by a Function Resume request.
> > > > > > > > > +     */
> > > > > > > > > +    if (f->func_suspended)
> > > > > > > > > +        return;
> > > > > > > > > +
> > > > > > > > > +    DBG(cdev, "ECM Resume\n");
> > > > > > > > > +
> > > > > > > > > +    gether_resume(&ecm->port);
> > > > > > > > > +}
> > > > > > > > > +
> > > > > > > > > +static int ecm_get_status(struct usb_function *f)
> > > > > > > > > +{
> > > > > > > > > +    struct usb_configuration *c = f->config;
> > > > > > > > > +
> > > > > > > > > +    /* D0 and D1 bit set to 0 if device is not wakeup capable */
> > > > > > > > > +    if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
> > > > > > > > > +        return 0;
> > > > > > > > > +
> > > > > > > > > +    return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
> > > > > > > > > +        USB_INTRF_STAT_FUNC_RW_CAP;
> > > > > > > > > +}
> > > > > > > > 
> > > > > > > > Why do we need to implement ecm_get_status if
> > > > > > > > it's already handled in
> > > > > > > > composite.c now?
> > > > > > > > 
> > > > > > > 
> > > > > > > Yes this can be removed now. Will modify accordingly.
> > > > > > > > > +
> > > > > > > > > +static int ecm_func_suspend(struct usb_function *f, u8 options)
> > > > > > > > > +{
> > > > > > > > > +    struct usb_composite_dev *cdev = f->config->cdev;
> > > > > > > > > +
> > > > > > > > > +    DBG(cdev, "func susp %u cmd\n", options);
> > > > > > > > > +
> > > > > > > > > +    if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
> > > > > > > > 
> > > > > > > > This feature selector doesn't indicate whether it's SetFeature or
> > > > > > > > ClearFeature request. ecm_func_suspend is supposed to be for
> > > > > > > > SetFeature(suspend) only. Perhaps we may have to
> > > > > > > > define func_resume()
> > > > > > > > for ClearFeature(suspend)?
> > > > > > > > 
> > > > > > 
> > > > > > > Host uses the same feature selector FUNCTION_SUSPEND
> > > > > > > for function suspend
> > > > > > > and function resume and func_suspend() callback can be used to
> > > > > > > handle both the cases ? The distinction comes whether it is a
> > > > > > 
> > > > > > How do you plan to handle that? Pass this info in some unused/reserved
> > > > > > bit of the "options" argument? Introduce a new parameter to the
> > > > > > func_suspend()?
> > > > > > 
> > > > > > If that's the case, then you need to update the document on
> > > > > > func_suspend() to also support ClearFeature(suspend). Right now it's
> > > > > > documented for SetFeature only. Also, make sure that other existing
> > > > > > function drivers will not break because of the change of the
> > > > > > func_suspend behavior.
> > > > > > 
> > > > > > > SetFeature(FUNCTION_SUSPEND) or
> > > > > > > ClearFeature(FUNCTION_SUSPEND) which can be
> > > > > > > easily done in the func_suspend callback itself. We
> > > > > > > can add another callback
> > > > > > > func_resume specific to
> > > > > > > ClearFeature(FUNCTION_SUSPEND) but wont that be
> > > > > > > redundant and more callback handling on function
> > > > > > > driver/composite side as
> > > > > > > well? Please let me know your opinion.
> > > > > > > 
> > > > > > 
> > > > > > We actually didn't properly define func_suspend and its
> > > > > > counter part. It
> > > > > > seems cleaner to me to introduce func_resume as it seems
> > > > > > more intuitive
> > > > > > and easier to read. Let me know how you plan to use func_suspend() for
> > > > > > both cases.
> > > > > > 
> > > > > 
> > > > > How about we handle function suspend resume in composite also? I mean
> > > > > something like this:
> > > > > 
> > > > > diff --git a/drivers/usb/gadget/composite.c
> > > > > b/drivers/usb/gadget/composite.c
> > > > > index 36add1879ed2..79dc055eb5f7 100644
> > > > > --- a/drivers/usb/gadget/composite.c
> > > > > +++ b/drivers/usb/gadget/composite.c
> > > > > @@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget
> > > > > *gadget, const struct usb_ctrlrequest *ctrl)
> > > > >            f = cdev->config->interface[intf];
> > > > >            if (!f)
> > > > >                break;
> > > > > -        status = f->get_status ? f->get_status(f) : 0;
> > > > > -        if (status < 0)
> > > > > -            break;
> > > > > +
> > > > > +        if (f->get_status) {
> > > > > +            status = f->get_status(f);
> > > > > +            if (status < 0)
> > > > > +                break;
> > > > > +        } else {
> > > > > +            if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
> > > > > +                status |= USB_INTRF_STAT_FUNC_RW_CAP;
> > > > > +                if (f->func_wakeup_armed)
> > > > > +                    status |= USB_INTRF_STAT_FUNC_RW;
> > > > > +            }
> > > > > +        }
> > > > >            put_unaligned_le16(status & 0x0000ffff, req->buf);
> > > > >            break;
> > > > >        /*
> > > > > @@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget
> > > > > *gadget, const struct usb_ctrlrequest *ctrl)
> > > > >                f = cdev->config->interface[intf];
> > > > >                if (!f)
> > > > >                    break;
> > > > > +            if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
> > > > > +                if (!(f->config->bmAttributes &
> > > > > USB_CONFIG_ATT_WAKEUP))
> > > > > +                    break;
> > > > > +
> > > > > +                f->func_wakeup_armed = (ctrl->bRequest ==
> > > > > USB_REQ_SET_FEATURE);
> > > > > +            }
> > > > > +
> > > > >                value = 0;
> > > > > -            if (f->func_suspend)
> > > > > +            if (f->func_suspend) {
> > > > >                    value = f->func_suspend(f, w_index >> 8);
> > > > > +            } else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
> > > > > +                if (f->suspend && && !f->func_suspended &&
> > > > > +                    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
> > > > > +                    f->suspend(f);
> > > > > +                    f->func_suspended = true;
> > > > > +                } else if (f->resume && f->func_suspended &&
> > > > > +                       ctrl->bRequest ==
> > > > > USB_REQ_CLEAR_FEATURE_FEATURE)) {
> 
> Linux based hosts, dont use ClearFeature(FUNCTION_SUSPEND) for function
> resume. They use SetFeature(FUNCTION_SUSPEND) itself but with Bit(0) i.e
> USB_INTRF_FUNC_SUSPEND_LP reset. So we may not be able to distinguish based
> on SET_FEATURE and CLEAR_FEATURE packets for function suspend and function
> resume. Instead we can generalize (w_index & USB_INTRF_FUNC_SUSPEND_LP) for
> function suspend and all other cases for function resume.
> 
> if (f->func_suspend) {
> 		value = f->func_suspend(f, w_index >> 8);
> } else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
> 		if (f->suspend && !f->func_suspended) {
> 				f->suspend(f);
> 				f->func_suspended = true;
> 		}
> } else {
> 		if (f->resume && f->func_suspended) {
> 				f->resume(f);
> 				f->func_suspended = false;
> 		}
> }
> 
> 

Ah.. right. That's possible also. Then can we check for both cases?
Something like this:

@ -1994,8 +2003,34 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
                        if (!f)
                                break;
                        value = 0;
-                       if (f->func_suspend)
+                       if (f->func_suspend) {
                                value = f->func_suspend(f, w_index >> 8);
+                       } else if (ctrl->bRequest == USB_REQ_SET_FEATURE) {
+                               if (!(f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) &&
+                                   (w_index & USB_INTRF_FUNC_SUSPEND_RW))
+                                               break;
+
+                               f->func_wakeup_armed = !!(w_index & USB_INTRF_FUNC_SUSPEND_RW);
+
+                               if ((w_index & USB_INTRF_FUNC_SUSPEND_LP) &&
+                                   f->suspend && !f->func_suspended) {
+                                       f->suspend(f);
+                                       f->func_suspended = true;
+                               } else if (f->resume && f->func_suspended) {
+                                       f->resume(f);
+                                       f->func_suspended = false;
+                               }
+                       } else if (ctrl->bRequest == USB_REQ_CLEAR_FEATURE_FEATURE)) {
+                               if (w_index & USB_INTRF_FUNC_SUSPEND_RW)
+                                       f->func_wakeup_armed = false;
+
+                               if ((w_index & USB_INTRF_FUNC_SUSPEND_LP) &&
+                                   f->resume && f->func_suspended) {
+                                       f->resume(f);
+                                       f->func_suspended = false;
+                               }
+                       }
+
                        if (value < 0) {
                                ERROR(cdev,
                                      "func_suspend() returned error %d\n",


Thanks,
Thinh

^ permalink raw reply	[flat|nested] 18+ messages in thread

* Re: [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support
  2023-03-24  0:59                   ` Thinh Nguyen
@ 2023-03-24  1:14                     ` Thinh Nguyen
  0 siblings, 0 replies; 18+ messages in thread
From: Thinh Nguyen @ 2023-03-24  1:14 UTC (permalink / raw)
  To: Elson Serrao; +Cc: gregkh, linux-kernel, linux-usb, quic_wcheng, quic_jackp

On Fri, Mar 24, 2023, Thinh Nguyen wrote:
> On Thu, Mar 23, 2023, Elson Serrao wrote:
> > 
> > 
> > On 3/20/2023 10:42 AM, Elson Serrao wrote:
> > > 
> > > 
> > > On 3/17/2023 7:26 PM, Thinh Nguyen wrote:
> > > > On Fri, Mar 17, 2023, Elson Serrao wrote:
> > > > > 
> > > > > 
> > > > > On 3/17/2023 4:20 PM, Thinh Nguyen wrote:
> > > > > > On Fri, Mar 17, 2023, Thinh Nguyen wrote:
> > > > > > > On Fri, Mar 17, 2023, Elson Serrao wrote:
> > > > > > > > 
> > > > > > > > 
> > > > > > > > On 3/16/2023 5:11 PM, Thinh Nguyen wrote:
> > > > > > > > > On Thu, Mar 16, 2023, Elson Roy Serrao wrote:
> > > > > > > > > > When host sends a suspend notification to the device, handle
> > > > > > > > > > the suspend callbacks in the function driver. Enhanced super
> > > > > > > > > > speed devices can support function suspend feature to put the
> > > > > > > > > > function in suspend state. Handle function suspend callback.
> > > > > > > > > > 
> > > > > > > > > > Depending on the remote wakeup capability the device can either
> > > > > > > > > > trigger a remote wakeup or wait for the host initiated resume to
> > > > > > > > > > start data transfer again.
> > > > > > > > > > 
> > > > > > > > > > Signed-off-by: Elson Roy Serrao <quic_eserrao@quicinc.com>
> > > > > > > > > > ---
> > > > > > > > > >     drivers/usb/gadget/function/f_ecm.c   |
> > > > > > > > > > 71 +++++++++++++++++++++++++++++++++++
> > > > > > > > > >     drivers/usb/gadget/function/u_ether.c |
> > > > > > > > > > 63 +++++++++++++++++++++++++++++++
> > > > > > > > > >     drivers/usb/gadget/function/u_ether.h |  4 ++
> > > > > > > > > >     3 files changed, 138 insertions(+)
> > > > > > > > > > 
> > > > > > > > > > diff --git
> > > > > > > > > > a/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > > b/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > > index a7ab30e..c43cd557 100644
> > > > > > > > > > --- a/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > > +++ b/drivers/usb/gadget/function/f_ecm.c
> > > > > > > > > > @@ -633,6 +633,8 @@ static void ecm_disable(struct usb_function *f)
> > > > > > > > > >         usb_ep_disable(ecm->notify);
> > > > > > > > > >         ecm->notify->desc = NULL;
> > > > > > > > > > +    f->func_suspended = false;
> > > > > > > > > > +    f->func_wakeup_armed = false;
> > > > > > > > > >     }
> > > > > > > > > >      /*-------------------------------------------------------------------------*/
> > > > > > > > > > 
> > > > > > > > > > @@ -885,6 +887,71 @@ static struct
> > > > > > > > > > usb_function_instance *ecm_alloc_inst(void)
> > > > > > > > > >         return &opts->func_inst;
> > > > > > > > > >     }
> > > > > > > > > > +static void ecm_suspend(struct usb_function *f)
> > > > > > > > > > +{
> > > > > > > > > > +    struct f_ecm *ecm = func_to_ecm(f);
> > > > > > > > > > +    struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > > > > > > > +
> > > > > > > > > > +    if (f->func_suspended) {
> > > > > > > > > > +        DBG(cdev, "Function already suspended\n");
> > > > > > > > > > +        return;
> > > > > > > > > > +    }
> > > > > > > > > > +
> > > > > > > > > > +    DBG(cdev, "ECM Suspend\n");
> > > > > > > > > > +
> > > > > > > > > > +    gether_suspend(&ecm->port);
> > > > > > > > > > +}
> > > > > > > > > > +
> > > > > > > > > > +static void ecm_resume(struct usb_function *f)
> > > > > > > > > > +{
> > > > > > > > > > +    struct f_ecm *ecm = func_to_ecm(f);
> > > > > > > > > > +    struct usb_composite_dev *cdev = ecm->port.func.config->cdev;
> > > > > > > > > > +
> > > > > > > > > > +    /*
> > > > > > > > > > +     * If the function is in USB3 Function
> > > > > > > > > > Suspend state, resume is
> > > > > > > > > > +     * canceled. In this case resume is
> > > > > > > > > > done by a Function Resume request.
> > > > > > > > > > +     */
> > > > > > > > > > +    if (f->func_suspended)
> > > > > > > > > > +        return;
> > > > > > > > > > +
> > > > > > > > > > +    DBG(cdev, "ECM Resume\n");
> > > > > > > > > > +
> > > > > > > > > > +    gether_resume(&ecm->port);
> > > > > > > > > > +}
> > > > > > > > > > +
> > > > > > > > > > +static int ecm_get_status(struct usb_function *f)
> > > > > > > > > > +{
> > > > > > > > > > +    struct usb_configuration *c = f->config;
> > > > > > > > > > +
> > > > > > > > > > +    /* D0 and D1 bit set to 0 if device is not wakeup capable */
> > > > > > > > > > +    if (!(USB_CONFIG_ATT_WAKEUP & c->bmAttributes))
> > > > > > > > > > +        return 0;
> > > > > > > > > > +
> > > > > > > > > > +    return (f->func_wakeup_armed ? USB_INTRF_STAT_FUNC_RW : 0) |
> > > > > > > > > > +        USB_INTRF_STAT_FUNC_RW_CAP;
> > > > > > > > > > +}
> > > > > > > > > 
> > > > > > > > > Why do we need to implement ecm_get_status if
> > > > > > > > > it's already handled in
> > > > > > > > > composite.c now?
> > > > > > > > > 
> > > > > > > > 
> > > > > > > > Yes this can be removed now. Will modify accordingly.
> > > > > > > > > > +
> > > > > > > > > > +static int ecm_func_suspend(struct usb_function *f, u8 options)
> > > > > > > > > > +{
> > > > > > > > > > +    struct usb_composite_dev *cdev = f->config->cdev;
> > > > > > > > > > +
> > > > > > > > > > +    DBG(cdev, "func susp %u cmd\n", options);
> > > > > > > > > > +
> > > > > > > > > > +    if (options & (USB_INTRF_FUNC_SUSPEND_LP >> 8)) {
> > > > > > > > > 
> > > > > > > > > This feature selector doesn't indicate whether it's SetFeature or
> > > > > > > > > ClearFeature request. ecm_func_suspend is supposed to be for
> > > > > > > > > SetFeature(suspend) only. Perhaps we may have to
> > > > > > > > > define func_resume()
> > > > > > > > > for ClearFeature(suspend)?
> > > > > > > > > 
> > > > > > > 
> > > > > > > > Host uses the same feature selector FUNCTION_SUSPEND
> > > > > > > > for function suspend
> > > > > > > > and function resume and func_suspend() callback can be used to
> > > > > > > > handle both the cases ? The distinction comes whether it is a
> > > > > > > 
> > > > > > > How do you plan to handle that? Pass this info in some unused/reserved
> > > > > > > bit of the "options" argument? Introduce a new parameter to the
> > > > > > > func_suspend()?
> > > > > > > 
> > > > > > > If that's the case, then you need to update the document on
> > > > > > > func_suspend() to also support ClearFeature(suspend). Right now it's
> > > > > > > documented for SetFeature only. Also, make sure that other existing
> > > > > > > function drivers will not break because of the change of the
> > > > > > > func_suspend behavior.
> > > > > > > 
> > > > > > > > SetFeature(FUNCTION_SUSPEND) or
> > > > > > > > ClearFeature(FUNCTION_SUSPEND) which can be
> > > > > > > > easily done in the func_suspend callback itself. We
> > > > > > > > can add another callback
> > > > > > > > func_resume specific to
> > > > > > > > ClearFeature(FUNCTION_SUSPEND) but wont that be
> > > > > > > > redundant and more callback handling on function
> > > > > > > > driver/composite side as
> > > > > > > > well? Please let me know your opinion.
> > > > > > > > 
> > > > > > > 
> > > > > > > We actually didn't properly define func_suspend and its
> > > > > > > counter part. It
> > > > > > > seems cleaner to me to introduce func_resume as it seems
> > > > > > > more intuitive
> > > > > > > and easier to read. Let me know how you plan to use func_suspend() for
> > > > > > > both cases.
> > > > > > > 
> > > > > > 
> > > > > > How about we handle function suspend resume in composite also? I mean
> > > > > > something like this:
> > > > > > 
> > > > > > diff --git a/drivers/usb/gadget/composite.c
> > > > > > b/drivers/usb/gadget/composite.c
> > > > > > index 36add1879ed2..79dc055eb5f7 100644
> > > > > > --- a/drivers/usb/gadget/composite.c
> > > > > > +++ b/drivers/usb/gadget/composite.c
> > > > > > @@ -1948,9 +1948,18 @@ composite_setup(struct usb_gadget
> > > > > > *gadget, const struct usb_ctrlrequest *ctrl)
> > > > > >            f = cdev->config->interface[intf];
> > > > > >            if (!f)
> > > > > >                break;
> > > > > > -        status = f->get_status ? f->get_status(f) : 0;
> > > > > > -        if (status < 0)
> > > > > > -            break;
> > > > > > +
> > > > > > +        if (f->get_status) {
> > > > > > +            status = f->get_status(f);
> > > > > > +            if (status < 0)
> > > > > > +                break;
> > > > > > +        } else {
> > > > > > +            if (f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) {
> > > > > > +                status |= USB_INTRF_STAT_FUNC_RW_CAP;
> > > > > > +                if (f->func_wakeup_armed)
> > > > > > +                    status |= USB_INTRF_STAT_FUNC_RW;
> > > > > > +            }
> > > > > > +        }
> > > > > >            put_unaligned_le16(status & 0x0000ffff, req->buf);
> > > > > >            break;
> > > > > >        /*
> > > > > > @@ -1971,9 +1980,28 @@ composite_setup(struct usb_gadget
> > > > > > *gadget, const struct usb_ctrlrequest *ctrl)
> > > > > >                f = cdev->config->interface[intf];
> > > > > >                if (!f)
> > > > > >                    break;
> > > > > > +            if (w_index & USB_INTRF_FUNC_SUSPEND_RW) {
> > > > > > +                if (!(f->config->bmAttributes &
> > > > > > USB_CONFIG_ATT_WAKEUP))
> > > > > > +                    break;
> > > > > > +
> > > > > > +                f->func_wakeup_armed = (ctrl->bRequest ==
> > > > > > USB_REQ_SET_FEATURE);
> > > > > > +            }
> > > > > > +
> > > > > >                value = 0;
> > > > > > -            if (f->func_suspend)
> > > > > > +            if (f->func_suspend) {
> > > > > >                    value = f->func_suspend(f, w_index >> 8);
> > > > > > +            } else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
> > > > > > +                if (f->suspend && && !f->func_suspended &&
> > > > > > +                    ctrl->bRequest == USB_REQ_SET_FEATURE)) {
> > > > > > +                    f->suspend(f);
> > > > > > +                    f->func_suspended = true;
> > > > > > +                } else if (f->resume && f->func_suspended &&
> > > > > > +                       ctrl->bRequest ==
> > > > > > USB_REQ_CLEAR_FEATURE_FEATURE)) {
> > 
> > Linux based hosts, dont use ClearFeature(FUNCTION_SUSPEND) for function
> > resume. They use SetFeature(FUNCTION_SUSPEND) itself but with Bit(0) i.e
> > USB_INTRF_FUNC_SUSPEND_LP reset. So we may not be able to distinguish based
> > on SET_FEATURE and CLEAR_FEATURE packets for function suspend and function
> > resume. Instead we can generalize (w_index & USB_INTRF_FUNC_SUSPEND_LP) for
> > function suspend and all other cases for function resume.
> > 
> > if (f->func_suspend) {
> > 		value = f->func_suspend(f, w_index >> 8);
> > } else if (w_index & USB_INTRF_FUNC_SUSPEND_LP) {
> > 		if (f->suspend && !f->func_suspended) {
> > 				f->suspend(f);
> > 				f->func_suspended = true;
> > 		}
> > } else {
> > 		if (f->resume && f->func_suspended) {
> > 				f->resume(f);
> > 				f->func_suspended = false;
> > 		}
> > }
> > 
> > 
> 
> Ah.. right. That's possible also. Then can we check for both cases?
> Something like this:
> 

Correction, ClearFeature clears the entire suspend option (remote wake +
suspend)

@@ -1994,8 +2003,32 @@ composite_setup(struct usb_gadget *gadget, const struct usb_ctrlrequest *ctrl)
                        if (!f)
                                break;
                        value = 0;
-                       if (f->func_suspend)
+                       if (f->func_suspend) {
                                value = f->func_suspend(f, w_index >> 8);
+                       } else if (ctrl->bRequest == USB_REQ_SET_FEATURE) {
+                               if (!(f->config->bmAttributes & USB_CONFIG_ATT_WAKEUP) &&
+                                   (w_index & USB_INTRF_FUNC_SUSPEND_RW))
+                                               break;
+
+                               f->func_wakeup_armed = !!(w_index & USB_INTRF_FUNC_SUSPEND_RW);
+
+                               if ((w_index & USB_INTRF_FUNC_SUSPEND_LP) &&
+                                   f->suspend && !f->func_suspended) {
+                                       f->suspend(f);
+                                       f->func_suspended = true;
+                               } else if (f->resume && f->func_suspended) {
+                                       f->resume(f);
+                                       f->func_suspended = false;
+                               }
+                       } else if (ctrl->bRequest == USB_REQ_CLEAR_FEATURE_FEATURE)) {
+                               f->func_wakeup_armed = false;
+
+                               if (f->resume && f->func_suspended) {
+                                       f->resume(f);
+                                       f->func_suspended = false;
+                               }
+                       }
+
                        if (value < 0) {
                                ERROR(cdev,
                                      "func_suspend() returned error %d\n",


Something like that.

Thanks,
Thinh

^ permalink raw reply	[flat|nested] 18+ messages in thread

end of thread, other threads:[~2023-03-24  1:14 UTC | newest]

Thread overview: 18+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-03-16 23:38 [PATCH v12 0/6] Add function suspend/resume and remote wakeup support Elson Roy Serrao
2023-03-16 23:38 ` [PATCH v12 1/6] usb: gadget: Properly configure the device for remote wakeup Elson Roy Serrao
2023-03-16 23:38 ` [PATCH v12 2/6] usb: dwc3: Add remote wakeup handling Elson Roy Serrao
2023-03-16 23:38 ` [PATCH v12 3/6] usb: gadget: Add function wakeup support Elson Roy Serrao
2023-03-16 23:38 ` [PATCH v12 4/6] usb: dwc3: Add function suspend and " Elson Roy Serrao
2023-03-16 23:38 ` [PATCH v12 5/6] usb: gadget: arm the function for triggering remote wakeup Elson Roy Serrao
2023-03-16 23:38 ` [PATCH v12 6/6] usb: gadget: f_ecm: Add suspend/resume and remote wakeup support Elson Roy Serrao
2023-03-17  0:11   ` Thinh Nguyen
2023-03-17 17:59     ` Elson Serrao
2023-03-17 21:28       ` Thinh Nguyen
2023-03-17 23:20         ` Thinh Nguyen
2023-03-18  0:23           ` Elson Serrao
2023-03-18  2:26             ` Thinh Nguyen
2023-03-20 17:42               ` Elson Serrao
2023-03-24  0:12                 ` Elson Serrao
2023-03-24  0:59                   ` Thinh Nguyen
2023-03-24  1:14                     ` Thinh Nguyen
2023-03-21  8:12   ` Dan Carpenter

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).