Skip to contentSkip to conversation

usb: hub: Cycle HUB power when initialization fails

RFC PATCH In review Mike Looijmans 3 participants 6 messages Nov 3, 2017 – Nov 9, 2017 USB

lore
Share & tools
lore
#0 lore raw

RFC PATCH 1/1 usb: hub: Cycle HUB power when initialization fails

Sometimes the USB device gets confused about the state of the initialization and the connection fails. In particular, the device thinks that it's already set up and running while the host thinks the device still needs to be configured. To work around this issue, power-cycle the hub's output to issue a sort of "reset" to the device. This makes the device restart its state machine and then the initialization succeeds.

This fixes problems where the kernel reports a list of errors like this:

usb 1-1.3: device not accepting address 19, error -71

The end result is a non-functioning device. After this patch, the sequence becomes like this:

usb 1-1.3: new high-speed USB device number 18 using ci_hdrc usb 1-1.3: device not accepting address 18, error -71 usb 1-1.3: new high-speed USB device number 19 using ci_hdrc usb 1-1.3: device not accepting address 19, error -71 usb 1-1-port3: attempt power cycle
usb 1-1.3: new high-speed USB device number 21 using ci_hdrc usb-storage 1-1.3:1.2: USB Mass Storage device detected

---
This is a fix I did for a customer which might be appropriate for upstream. What do you think?

 drivers/usb/core/hub.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)
drivers/usb/core/hub.c +10 -3
index e9ce6bb..a30c1e7 100644
@@ -2611,7 +2611,7 @@ static unsigned hub_is_wusb(struct usb_hub *hub)
 #define PORT_RESET_TRIES	5
 #define SET_ADDRESS_TRIES	2
 #define GET_DESCRIPTOR_TRIES	2
-#define SET_CONFIG_TRIES	(2 * (use_both_schemes + 1))
+#define SET_CONFIG_TRIES	(4 * (use_both_schemes + 1))
 #define USE_NEW_SCHEME(i)	((i) / 2 == (int)old_scheme_first)
 
 #define HUB_ROOT_RESET_TIME	60	/* times are in msec */
@@ -4805,7 +4805,6 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
 
 	status = 0;
 	for (i = 0; i < SET_CONFIG_TRIES; i++) {
-
 		/* reallocate for each attempt, since references
 		 * to the previous one can escape in various ways
 		 */
@@ -4935,6 +4934,15 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
 		usb_put_dev(udev);
 		if ((status == -ENOTCONN) || (status == -ENOTSUPP))
 			break;
+
+		/* When halfway through our retry count, power-cycle the port */
+		if (i == (SET_CONFIG_TRIES / 2) - 1) {
+			dev_info(&port_dev->dev, "attempt power cycle\n");
+			usb_hub_set_port_power(hdev, hub, port1, false);
+			msleep(800);
+			usb_hub_set_port_power(hdev, hub, port1, true);
+			msleep(hub_power_on_good_delay(hub));
+		}
 	}
 	if (hub->hdev->parent ||
 			!hcd->driver->port_handed_over ||
@@ -5476,7 +5484,6 @@ static int usb_reset_and_verify_device(struct usb_device *udev)
 	udev->bos = NULL;
 
 	for (i = 0; i < SET_CONFIG_TRIES; ++i) {
-
 		/* ep0 maxpacket size may change; let the HCD know about it.
 		 * Other endpoints will be handled by re-enumeration. */
 		usb_ep0_reinit(udev);
Signature · 1 line
1.9.1
#1 lore raw
Quoted 3 lines · Mike Looijmans

Sometimes the USB device gets confused about the state of the initialization and the connection fails. In particular, the device thinks that it's already set up and running while the host thinks the device still needs to be configured. To

How do you know that this is really the issue? How can the device think it's already set if it doesn't have an assigned address?

Quoted 31 lines

work around this issue, power-cycle the hub's output to issue a sort of "reset" to the device. This makes the device restart its state machine and then the initialization succeeds.

This fixes problems where the kernel reports a list of errors like this:

usb 1-1.3: device not accepting address 19, error -71

The end result is a non-functioning device. After this patch, the sequence becomes like this:

usb 1-1.3: new high-speed USB device number 18 using ci_hdrc usb 1-1.3: device not accepting address 18, error -71 usb 1-1.3: new high-speed USB device number 19 using ci_hdrc usb 1-1.3: device not accepting address 19, error -71 usb 1-1-port3: attempt power cycle
usb 1-1.3: new high-speed USB device number 21 using ci_hdrc usb-storage 1-1.3:1.2: USB Mass Storage device detected

---
This is a fix I did for a customer which might be appropriate for upstream. What do you think?

 drivers/usb/core/hub.c | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)
drivers/usb/core/hub.c +0 -0
index e9ce6bb..a30c1e7 100644

drivers/usb/core/hub.c static unsigned hub_is_wusb(struct usb_hub *hub) · line 2614

Quoted 5 lines
drivers/usb/core/hub.c +1 -1
@@ -2611,7 +2611,7 @@ static unsigned hub_is_wusb(struct usb_hub *hub)
 #define PORT_RESET_TRIES	5
 #define SET_ADDRESS_TRIES	2
 #define GET_DESCRIPTOR_TRIES	2
-#define SET_CONFIG_TRIES	(2 * (use_both_schemes + 1))
+#define SET_CONFIG_TRIES	(4 * (use_both_schemes + 1))

We already have too many retry loops. I am not keen on the idea of adding even more. How about leaving this value the same and adding the power cycle?

The ideal, however, would be to find out what is wrong with the device and see what needs to be done to fix it properly. This change won't work on many computers (desktops and laptops) because they don't have real USB port-power switching. A lot of hubs don't have it either.

drivers/usb/core/hub.c static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus, · line 4808

Quoted 8 lines
drivers/usb/core/hub.c +0 -1
 #define USE_NEW_SCHEME(i)	((i) / 2 == (int)old_scheme_first)
 
 #define HUB_ROOT_RESET_TIME	60	/* times are in msec */
@@ -4805,7 +4805,6 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
 
 	status = 0;
 	for (i = 0; i < SET_CONFIG_TRIES; i++) {
-

Gratuitous whitespace change.

Quoted 20 lines
drivers/usb/core/hub.c +9 -0
 		/* reallocate for each attempt, since references
 		 * to the previous one can escape in various ways
 		 */
@@ -4935,6 +4934,15 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
 		usb_put_dev(udev);
 		if ((status == -ENOTCONN) || (status == -ENOTSUPP))
 			break;
+
+		/* When halfway through our retry count, power-cycle the port */
+		if (i == (SET_CONFIG_TRIES / 2) - 1) {
+			dev_info(&port_dev->dev, "attempt power cycle\n");
+			usb_hub_set_port_power(hdev, hub, port1, false);
+			msleep(800);
+			usb_hub_set_port_power(hdev, hub, port1, true);
+			msleep(hub_power_on_good_delay(hub));
+		}
 	}
 	if (hub->hdev->parent ||
 			!hcd->driver->port_handed_over ||

drivers/usb/core/hub.c static int usb_reset_and_verify_device(struct usb_device *udev) · line 5479

Quoted 4 lines
drivers/usb/core/hub.c +0 -1
@@ -5476,7 +5484,6 @@ static int usb_reset_and_verify_device(struct usb_device *udev)
 	udev->bos = NULL;
 
 	for (i = 0; i < SET_CONFIG_TRIES; ++i) {
-

Another gratuitous change.

drivers/usb/core/hub.c static int usb_reset_and_verify_device(struct usb_device *udev) · line 5489

Quoted 3 lines
drivers/usb/core/hub.c +0 -0
 		/* ep0 maxpacket size may change; let the HCD know about it.
 		 * Other endpoints will be handled by re-enumeration. */
 		usb_ep0_reinit(udev);

Alan Stern

#2 lore raw
Quoted 8 lines · Alan Stern
Quoted 3 lines · Mike Looijmans

Sometimes the USB device gets confused about the state of the initialization and the connection fails. In particular, the device thinks that it's already set up and running while the host thinks the device still needs to be configured. To

How do you know that this is really the issue? How can the device think it's already set if it doesn't have an assigned address?

It seems to me that the device just doesn't react at all on the host requests to assign an address. I've seen this happen with various custom mass-storage like appliances, but also DVB tuners and such. The device won't return to a working state until you unplug it and put it back, or, and that's what the patch does, just power-cycle the USB port, which has the same effect.

Quoted 26 lines
Quoted 26 lines

work around this issue, power-cycle the hub's output to issue a sort of "reset" to the device. This makes the device restart its state machine and then the initialization succeeds.

This fixes problems where the kernel reports a list of errors like this:

usb 1-1.3: device not accepting address 19, error -71

The end result is a non-functioning device. After this patch, the sequence becomes like this:

usb 1-1.3: new high-speed USB device number 18 using ci_hdrc usb 1-1.3: device not accepting address 18, error -71 usb 1-1.3: new high-speed USB device number 19 using ci_hdrc usb 1-1.3: device not accepting address 19, error -71 usb 1-1-port3: attempt power cycle
usb 1-1.3: new high-speed USB device number 21 using ci_hdrc usb-storage 1-1.3:1.2: USB Mass Storage device detected

---
This is a fix I did for a customer which might be appropriate for upstream. What do you think?

  drivers/usb/core/hub.c | 13 ++++++++++---
  1 file changed, 10 insertions(+), 3 deletions(-)
Quoted 14 lines

drivers/usb/core/hub.c static unsigned hub_is_wusb(struct usb_hub *hub) · line 2614

Quoted 10 lines
drivers/usb/core/hub.c +1 -1
index e9ce6bb..a30c1e7 100644
@@ -2611,7 +2611,7 @@ static unsigned hub_is_wusb(struct usb_hub *hub)
  #define PORT_RESET_TRIES	5
  #define SET_ADDRESS_TRIES	2
  #define GET_DESCRIPTOR_TRIES	2
-#define SET_CONFIG_TRIES	(2 * (use_both_schemes + 1))
+#define SET_CONFIG_TRIES	(4 * (use_both_schemes + 1))

We already have too many retry loops. I am not keen on the idea of adding even more. How about leaving this value the same and adding the power cycle?

I'm fine with that as well.

Quoted 4 lines

The ideal, however, would be to find out what is wrong with the device and see what needs to be done to fix it properly. This change won't work on many computers (desktops and laptops) because they don't have real USB port-power switching. A lot of hubs don't have it either.

I'm pretty sure it's the device's fault, but they're out there and probably not upgradable anyway if you could get the manufacturer to pick up the phone.

On desktop/laptop machines the problem isn't as pressing since there's a often a user there who can unplug the thing. It's really nasty on embedded systems, and they tend to have the USB power wired through a supply limiter/switch. I'm thinking worst that could happen on desktops is that the patch won't have any effect at all.

So what do you think, is this worth a v2 patch for general consumption or should I keep this a "private" patch for systems that have demonstrated to benefit from it?

Quoted 44 lines

drivers/usb/core/hub.c static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus, · line 4808

Quoted 8 lines
drivers/usb/core/hub.c +0 -1
  #define USE_NEW_SCHEME(i)	((i) / 2 == (int)old_scheme_first)
 
  #define HUB_ROOT_RESET_TIME	60	/* times are in msec */
@@ -4805,7 +4805,6 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
 
  	status = 0;
  	for (i = 0; i < SET_CONFIG_TRIES; i++) {
-

Gratuitous whitespace change.

Quoted 20 lines
drivers/usb/core/hub.c +9 -0
  		/* reallocate for each attempt, since references
  		 * to the previous one can escape in various ways
  		 */
@@ -4935,6 +4934,15 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
  		usb_put_dev(udev);
  		if ((status == -ENOTCONN) || (status == -ENOTSUPP))
  			break;
+
+		/* When halfway through our retry count, power-cycle the port */
+		if (i == (SET_CONFIG_TRIES / 2) - 1) {
+			dev_info(&port_dev->dev, "attempt power cycle\n");
+			usb_hub_set_port_power(hdev, hub, port1, false);
+			msleep(800);
+			usb_hub_set_port_power(hdev, hub, port1, true);
+			msleep(hub_power_on_good_delay(hub));
+		}
  	}
  	if (hub->hdev->parent ||
  			!hcd->driver->port_handed_over ||

drivers/usb/core/hub.c static int usb_reset_and_verify_device(struct usb_device *udev) · line 5479

Quoted 4 lines
drivers/usb/core/hub.c +0 -1
@@ -5476,7 +5484,6 @@ static int usb_reset_and_verify_device(struct usb_device *udev)
  	udev->bos = NULL;
 
  	for (i = 0; i < SET_CONFIG_TRIES; ++i) {
-

Another gratuitous change.

drivers/usb/core/hub.c static int usb_reset_and_verify_device(struct usb_device *udev) · line 5489

Quoted 3 lines
drivers/usb/core/hub.c +0 -0
  		/* ep0 maxpacket size may change; let the HCD know about it.
  		 * Other endpoints will be handled by re-enumeration. */
  		usb_ep0_reinit(udev);

Alan Stern

Signature · 16 lines
Mike Looijmans


Kind regards,

Mike Looijmans
System Expert

TOPIC Products
Materiaalweg 4, NL-5681 RJ Best
Postbus 440, NL-5680 AK Best
Telefoon: +31 (0) 499 33 69 79
E-mail: [email protected]
Website: www.topicproducts.com

Please consider the environment before printing this e-mail
#3 lore raw
Quoted 12 lines · Mike Looijmans
Quoted 8 lines · Alan Stern
Quoted 3 lines · Mike Looijmans

Sometimes the USB device gets confused about the state of the initialization and the connection fails. In particular, the device thinks that it's already set up and running while the host thinks the device still needs to be configured. To

How do you know that this is really the issue?  How can the device think it's already set if it doesn't have an assigned address?

It seems to me that the device just doesn't react at all on the host requests to assign an address. I've seen this happen with various custom mass-storage like appliances, but also DVB tuners and such. The device won't return to a working state until you unplug it and put it back, or, and that's what the patch does, just power-cycle the USB port, which has the same effect.

We have seen this problem with Ethernet dongles left in a bad/hung state during a previous boot, so it definitely does happen. This happened, for example, with the r8152 driver with upstream commit 2f25abe6bac ("r8152: prevent the driver from transmitting packets with carrier off") missing.

Problem though, as mentioned, is that many hubs don't implement this feature, and if I understand correctly root hubs don't implement it either.

Guenter

Quoted 99 lines
Quoted 26 lines
Quoted 26 lines

work around this issue, power-cycle the hub's output to issue a sort of "reset" to the device. This makes the device restart its state machine and then the initialization succeeds.

This fixes problems where the kernel reports a list of errors like this:

usb 1-1.3: device not accepting address 19, error -71

The end result is a non-functioning device. After this patch, the sequence becomes like this:

usb 1-1.3: new high-speed USB device number 18 using ci_hdrc usb 1-1.3: device not accepting address 18, error -71 usb 1-1.3: new high-speed USB device number 19 using ci_hdrc usb 1-1.3: device not accepting address 19, error -71 usb 1-1-port3: attempt power cycle
usb 1-1.3: new high-speed USB device number 21 using ci_hdrc usb-storage 1-1.3:1.2: USB Mass Storage device detected

---
This is a fix I did for a customer which might be appropriate for upstream. What do you think?

  drivers/usb/core/hub.c | 13 ++++++++++---

  1 file changed, 10 insertions(+), 3 deletions(-)
Quoted 14 lines

drivers/usb/core/hub.c static unsigned hub_is_wusb(struct usb_hub *hub) · line 2614

Quoted 10 lines
drivers/usb/core/hub.c +1 -1
index e9ce6bb..a30c1e7 100644
@@ -2611,7 +2611,7 @@ static unsigned hub_is_wusb(struct usb_hub *hub)
   #define PORT_RESET_TRIES    5
   #define SET_ADDRESS_TRIES    2
   #define GET_DESCRIPTOR_TRIES    2
-#define SET_CONFIG_TRIES    (2 * (use_both_schemes + 1))
+#define SET_CONFIG_TRIES    (4 * (use_both_schemes + 1))

We already have too many retry loops.  I am not keen on the idea of adding even more.  How about leaving this value the same and adding the power cycle?

I'm fine with that as well.

Quoted 4 lines

The ideal, however, would be to find out what is wrong with the device and see what needs to be done to fix it properly.  This change won't work on many computers (desktops and laptops) because they don't have real USB port-power switching.  A lot of hubs don't have it either.

I'm pretty sure it's the device's fault, but they're out there and probably not upgradable anyway if you could get the manufacturer to pick up the phone.

On desktop/laptop machines the problem isn't as pressing since there's a often a user there who can unplug the thing. It's really nasty on embedded systems, and they tend to have the USB power wired through a supply limiter/switch. I'm thinking worst that could happen on desktops is that the patch won't have any effect at all.

So what do you think, is this worth a v2 patch for general consumption or should I keep this a "private" patch for systems that have demonstrated to benefit from it?

Quoted 41 lines

drivers/usb/core/hub.c static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus, · line 4807

Quoted 6 lines
drivers/usb/core/hub.c +0 -1
   #define USE_NEW_SCHEME(i)    ((i) / 2 == (int)old_scheme_first)
   #define HUB_ROOT_RESET_TIME    60    /* times are in msec */
@@ -4805,7 +4805,6 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
       status = 0;
       for (i = 0; i < SET_CONFIG_TRIES; i++) {
-

Gratuitous whitespace change.

Quoted 20 lines
drivers/usb/core/hub.c +9 -0
           /* reallocate for each attempt, since references
            * to the previous one can escape in various ways
            */
@@ -4935,6 +4934,15 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
           usb_put_dev(udev);
           if ((status == -ENOTCONN) || (status == -ENOTSUPP))
               break;
+
+        /* When halfway through our retry count, power-cycle the port */
+        if (i == (SET_CONFIG_TRIES / 2) - 1) {
+            dev_info(&port_dev->dev, "attempt power cycle\n");
+            usb_hub_set_port_power(hdev, hub, port1, false);
+            msleep(800);
+            usb_hub_set_port_power(hdev, hub, port1, true);
+            msleep(hub_power_on_good_delay(hub));
+        }
       }

      if (hub->hdev->parent ||
              !hcd->driver->port_handed_over ||

drivers/usb/core/hub.c static int usb_reset_and_verify_device(struct usb_device *udev) · line 5478

Quoted 3 lines
drivers/usb/core/hub.c +0 -1
@@ -5476,7 +5484,6 @@ static int usb_reset_and_verify_device(struct usb_device *udev)
       udev->bos = NULL;
       for (i = 0; i < SET_CONFIG_TRIES; ++i) {
-

Another gratuitous change.

drivers/usb/core/hub.c static int usb_reset_and_verify_device(struct usb_device *udev) · line 5489

Quoted 3 lines
drivers/usb/core/hub.c +0 -0
           /* ep0 maxpacket size may change; let the HCD know about it.
            * Other endpoints will be handled by re-enumeration. */
           usb_ep0_reinit(udev);

Alan Stern

#4 lore raw
In reply to Mike Looijmans
Quoted 76 lines · Mike Looijmans
Quoted 8 lines · Alan Stern
Quoted 3 lines · Mike Looijmans

Sometimes the USB device gets confused about the state of the initialization and the connection fails. In particular, the device thinks that it's already set up and running while the host thinks the device still needs to be configured. To

How do you know that this is really the issue? How can the device think it's already set if it doesn't have an assigned address?

It seems to me that the device just doesn't react at all on the host requests to assign an address. I've seen this happen with various custom mass-storage like appliances, but also DVB tuners and such. The device won't return to a working state until you unplug it and put it back, or, and that's what the patch does, just power-cycle the USB port, which has the same effect.

Quoted 26 lines
Quoted 26 lines

work around this issue, power-cycle the hub's output to issue a sort of "reset" to the device. This makes the device restart its state machine and then the initialization succeeds.

This fixes problems where the kernel reports a list of errors like this:

usb 1-1.3: device not accepting address 19, error -71

The end result is a non-functioning device. After this patch, the sequence becomes like this:

usb 1-1.3: new high-speed USB device number 18 using ci_hdrc usb 1-1.3: device not accepting address 18, error -71 usb 1-1.3: new high-speed USB device number 19 using ci_hdrc usb 1-1.3: device not accepting address 19, error -71 usb 1-1-port3: attempt power cycle
usb 1-1.3: new high-speed USB device number 21 using ci_hdrc usb-storage 1-1.3:1.2: USB Mass Storage device detected

---
This is a fix I did for a customer which might be appropriate for upstream. What do you think?

  drivers/usb/core/hub.c | 13 ++++++++++---
  1 file changed, 10 insertions(+), 3 deletions(-)
Quoted 14 lines

drivers/usb/core/hub.c static unsigned hub_is_wusb(struct usb_hub *hub) · line 2614

Quoted 10 lines
drivers/usb/core/hub.c +1 -1
index e9ce6bb..a30c1e7 100644
@@ -2611,7 +2611,7 @@ static unsigned hub_is_wusb(struct usb_hub *hub)
  #define PORT_RESET_TRIES	5
  #define SET_ADDRESS_TRIES	2
  #define GET_DESCRIPTOR_TRIES	2
-#define SET_CONFIG_TRIES	(2 * (use_both_schemes + 1))
+#define SET_CONFIG_TRIES	(4 * (use_both_schemes + 1))

We already have too many retry loops. I am not keen on the idea of adding even more. How about leaving this value the same and adding the power cycle?

I'm fine with that as well.

Quoted 4 lines

The ideal, however, would be to find out what is wrong with the device and see what needs to be done to fix it properly. This change won't work on many computers (desktops and laptops) because they don't have real USB port-power switching. A lot of hubs don't have it either.

I'm pretty sure it's the device's fault, but they're out there and probably not upgradable anyway if you could get the manufacturer to pick up the phone.

On desktop/laptop machines the problem isn't as pressing since there's a often a user there who can unplug the thing. It's really nasty on embedded systems, and they tend to have the USB power wired through a supply limiter/switch.
I'm thinking worst that could happen on desktops is that the patch won't have any effect at all.

Quoted 3 lines

So what do you think, is this worth a v2 patch for general consumption or should I keep this a "private" patch for systems that have demonstrated to benefit from it?

We might as well apply a v2 patch, since it will help on some systems.

Alan stern

Acked-by
#5 lore raw
In reply to Alan Stern
Quoted 42 lines · Mike Looijmans

Sometimes the USB device gets confused about the state of the initialization and the connection fails. In particular, the device thinks that it's already set up and running while the host thinks the device still needs to be configured. To work around this issue, power-cycle the hub's output to issue a sort of "reset" to the device. This makes the device restart its state machine and then the initialization succeeds.

This fixes problems where the kernel reports a list of errors like this:

usb 1-1.3: device not accepting address 19, error -71

The end result is a non-functioning device. After this patch, the sequence becomes like this:

usb 1-1.3: new high-speed USB device number 18 using ci_hdrc usb 1-1.3: device not accepting address 18, error -71 usb 1-1.3: new high-speed USB device number 19 using ci_hdrc usb 1-1.3: device not accepting address 19, error -71 usb 1-1-port3: attempt power cycle
usb 1-1.3: new high-speed USB device number 21 using ci_hdrc usb-storage 1-1.3:1.2: USB Mass Storage device detected

---
v2: On feedback from Alan Stern:
No blank line changes
Don't further increase the retry count
Delay 2x the power-up delay (100ms minimum) instead of magic "800"

 drivers/usb/core/hub.c | 9 +++++++++
 1 file changed, 9 insertions(+)
drivers/usb/core/hub.c +2 -0
index e9ce6bb..8f7d942 100644
@@ -4935,6 +4935,15 @@ static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus,
 		usb_put_dev(udev);
 		if ((status == -ENOTCONN) || (status == -ENOTSUPP))
 			break;
+
+		/* When halfway through our retry count, power-cycle the port */

drivers/usb/core/hub.c static void hub_port_connect(struct usb_hub *hub, int port1, u16 portstatus, · line 4949

Quoted 10 lines
drivers/usb/core/hub.c +7 -0
+		if (i == (SET_CONFIG_TRIES / 2) - 1) {
+			dev_info(&port_dev->dev, "attempt power cycle\n");
+			usb_hub_set_port_power(hdev, hub, port1, false);
+			msleep(2 * hub_power_on_good_delay(hub));
+			usb_hub_set_port_power(hdev, hub, port1, true);
+			msleep(hub_power_on_good_delay(hub));
+		}
 	}
 	if (hub->hdev->parent ||
 			!hcd->driver->port_handed_over ||