Skip to contentSkip to conversation

fixes for hangs and error reporting in CDC_WDM

RFC In review Oliver Neukum 3 participants 17 messages Sep 22, 2020 – Sep 29, 2020 USB

lore
Share & tools
lore
#0 lore raw

RFC 1/1 fixes for hangs and error reporting in CDC_WDM

Stress testing has shown that CDC-WDM has some issues with hangs and error reporting

1. wakeups are not correctly handled in multhreaded environments
2. unresponsive hardware is not handled
3. errors are not correctly reported. This needs flush() to be implemented.

This version makes wdm_flush() use interruptible sleep.

For easier review all squashed together:

drivers/usb/class/cdc-wdm.c +55 -8
@@ -58,6 +58,9 @@ MODULE_DEVICE_TABLE (usb, wdm_ids);
 
 #define WDM_MAX                        16
 
+/* flush() is uninterruptible, but we cannot wait forever */
+#define WDM_FLUSH_TIMEOUT      (30 * HZ)
+
 /* CDC-WMC r1.1 requires wMaxCommand to be "at least 256 decimal (0x100)" */
 #define WDM_DEFAULT_BUFSIZE    256
 
@@ -151,7 +154,7 @@ static void wdm_out_callback(struct urb *urb)
        kfree(desc->outbuf);
        desc->outbuf = NULL;
        clear_bit(WDM_IN_USE, &desc->flags);
-       wake_up(&desc->wait);
+       wake_up_all(&desc->wait);
 }
 
 static void wdm_in_callback(struct urb *urb)
@@ -393,6 +396,9 @@ static ssize_t wdm_write
        if (test_bit(WDM_RESETTING, &desc->flags))
                r = -EIO;
 
+       if (test_bit(WDM_DISCONNECTING, &desc->flags))
+               r = -ENODEV;
+
        if (r < 0) {
                rv = r;
                goto out_free_mem_pm;
@@ -424,6 +430,7 @@ static ssize_t wdm_write
        if (rv < 0) {
                desc->outbuf = NULL;
                clear_bit(WDM_IN_USE, &desc->flags);
+               wake_up_all(&desc->wait); /* for flush() */
                dev_err(&desc->intf->dev, "Tx URB error: %d\n", rv);
                rv = usb_translate_errors(rv);
                goto out_free_mem_pm;
@@ -583,11 +590,39 @@ static ssize_t wdm_read
        return rv;
 }
 
+/*
+ * The difference to flush is that we wait forever. If you don't like
+ * that behavior, you need to send a signal.
+ */
+
+static int wdm_fsync(struct file *file, loff_t start, loff_t end, int datasync)
+{
+       struct wdm_device *desc = file->private_data;
+       int rv;
+
+       rv = wait_event_interruptible(desc->wait,
+                       !test_bit(WDM_IN_USE, &desc->flags) ||
+                       test_bit(WDM_DISCONNECTING, &desc->flags));
+
+       if (test_bit(WDM_DISCONNECTING, &desc->flags))
+               return -ENODEV;
+       if (rv < 0)
+               return -EINTR;
+
+       spin_lock_irq(&desc->iuspin);
+       rv = desc->werr;
+       desc->werr = 0;
+       spin_unlock_irq(&desc->iuspin);
+
+       return usb_translate_errors(rv);
+}
+
 static int wdm_flush(struct file *file, fl_owner_t id)
 {
        struct wdm_device *desc = file->private_data;
+       int rv;
 
-       wait_event(desc->wait,
+       rv = wait_event_interruptible_timeout(desc->wait,
                        /*
                         * needs both flags. We cannot do with one
                         * because resetting it would cause a race
@@ -595,16 +630,27 @@ static int wdm_flush(struct file *file, fl_owner_t id)
                         * a disconnect
                         */
                        !test_bit(WDM_IN_USE, &desc->flags) ||
-                       test_bit(WDM_DISCONNECTING, &desc->flags));
+                       test_bit(WDM_DISCONNECTING, &desc->flags),
+                       WDM_FLUSH_TIMEOUT);
 
-       /* cannot dereference desc->intf if WDM_DISCONNECTING */
+       /*
+        * to report the correct error.
+        * This is best effort
+        * We are inevitably racing with the hardware.
+        */
        if (test_bit(WDM_DISCONNECTING, &desc->flags))
                return -ENODEV;
-       if (desc->werr < 0)
-               dev_err(&desc->intf->dev, "Error in flush path: %d\n",
-                       desc->werr);
+       if (!rv)
+               return -EIO;
+       if (rv < 0)
+               return -EINTR;
+
+       spin_lock_irq(&desc->iuspin);
+       rv = desc->werr;
+       desc->werr = 0;
+       spin_unlock_irq(&desc->iuspin);
 
-       return usb_translate_errors(desc->werr);
+       return usb_translate_errors(rv);
 }
 
 static __poll_t wdm_poll(struct file *file, struct poll_table_struct *wait)
@@ -729,6 +775,7 @@ static const struct file_operations wdm_fops = {
        .owner =        THIS_MODULE,
        .read =         wdm_read,
        .write =        wdm_write,
+       .fsync =        wdm_fsync,
        .open =         wdm_open,
        .flush =        wdm_flush,
        .release =      wdm_release,
#1 lore raw

RFC 1/7 CDC-WDM: fix hangs in flush() in multithreaded cases

In a multithreaded scenario an arbitrary number of threads can be in wdm_write() and in wdm_flush(). Hence whenever WDM_IN_USE is reset, all possible waiters need to be notified. This is true whether this is due to IO completing or to an error starting IO.

---
 drivers/usb/class/cdc-wdm.c | 11 +++++++----
 1 file changed, 7 insertions(+), 4 deletions(-)
drivers/usb/class/cdc-wdm.c +7 -4
index e3db6fbeadef..adb3fc307083 100644
@@ -151,7 +151,7 @@ static void wdm_out_callback(struct urb *urb)
 	kfree(desc->outbuf);
 	desc->outbuf = NULL;
 	clear_bit(WDM_IN_USE, &desc->flags);
-	wake_up(&desc->wait);
+	wake_up_all(&desc->wait);
 }
 
 static void wdm_in_callback(struct urb *urb)
@@ -424,6 +424,7 @@ static ssize_t wdm_write
 	if (rv < 0) {
 		desc->outbuf = NULL;
 		clear_bit(WDM_IN_USE, &desc->flags);
+		wake_up_all(&desc->wait); /* for flush() */
 		dev_err(&desc->intf->dev, "Tx URB error: %d\n", rv);
 		rv = usb_translate_errors(rv);
 		goto out_free_mem_pm;
@@ -586,6 +587,7 @@ static ssize_t wdm_read
 static int wdm_flush(struct file *file, fl_owner_t id)
 {
 	struct wdm_device *desc = file->private_data;
+	int rv;
 
 	wait_event(desc->wait,
 			/*
@@ -600,11 +602,12 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 	/* cannot dereference desc->intf if WDM_DISCONNECTING */
 	if (test_bit(WDM_DISCONNECTING, &desc->flags))
 		return -ENODEV;
-	if (desc->werr < 0)
+	rv = desc->werr;
+	if (rv < 0)
 		dev_err(&desc->intf->dev, "Error in flush path: %d\n",
-			desc->werr);
+			rv);
 
-	return usb_translate_errors(desc->werr);
+	return usb_translate_errors(rv);
 }
 
 static __poll_t wdm_poll(struct file *file, struct poll_table_struct *wait)
Signature · 1 line
2.16.4
#2 lore raw

RFC 2/7 CDC-WDM: introduce a timeout in flush()

In reply to Oliver Neukum

Malicious or defective hardware may cease communication while flush() is running. In last consequence we need a timeout, as hardware that remains silent must be ignored.

Making the wait for IO interruptible would not solve the issue. While it would avoid a hang, it would not allow any progress and we would end up with an unclosable fd.

The 30 seconds are coming out of thin air.

---
 drivers/usb/class/cdc-wdm.c | 11 +++++++++--
 1 file changed, 9 insertions(+), 2 deletions(-)
drivers/usb/class/cdc-wdm.c +9 -2
index adb3fc307083..b3a3f249a915 100644
@@ -58,6 +58,9 @@ MODULE_DEVICE_TABLE (usb, wdm_ids);
 
 #define WDM_MAX			16
 
+/* flush() is uninterruptible, but we cannot wait forever */
+#define WDM_FLUSH_TIMEOUT	(30 * HZ)
+
 /* CDC-WMC r1.1 requires wMaxCommand to be "at least 256 decimal (0x100)" */
 #define WDM_DEFAULT_BUFSIZE	256
 
@@ -589,7 +592,7 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 	struct wdm_device *desc = file->private_data;
 	int rv;
 
-	wait_event(desc->wait,
+	rv = wait_event_timeout(desc->wait,
 			/*
 			 * needs both flags. We cannot do with one
 			 * because resetting it would cause a race
@@ -597,11 +600,15 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 			 * a disconnect
 			 */
 			!test_bit(WDM_IN_USE, &desc->flags) ||
-			test_bit(WDM_DISCONNECTING, &desc->flags));
+			test_bit(WDM_DISCONNECTING, &desc->flags),
+			WDM_FLUSH_TIMEOUT);
 
 	/* cannot dereference desc->intf if WDM_DISCONNECTING */
 	if (test_bit(WDM_DISCONNECTING, &desc->flags))
 		return -ENODEV;
+	if (!rv)
+		return -EIO;
+
 	rv = desc->werr;
 	if (rv < 0)
 		dev_err(&desc->intf->dev, "Error in flush path: %d\n",
Signature · 1 line
2.16.4
#3 lore raw

RFC 3/7 CDC-WDM: remove use of intf->dev after potential disconnect

In reply to Oliver Neukum

After a disconnect intf->dev is not a valid pointer any longer. As flush() uses it only for logging purposes logging is not worth it. Remove the dev_err()

---
 drivers/usb/class/cdc-wdm.c | 9 +++++----
 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/usb/class/cdc-wdm.c +5 -4
index b3a3f249a915..1ca2c85a977d 100644
@@ -603,16 +603,17 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 			test_bit(WDM_DISCONNECTING, &desc->flags),
 			WDM_FLUSH_TIMEOUT);
 
-	/* cannot dereference desc->intf if WDM_DISCONNECTING */
+	/*
+	 * to report the correct error.
+	 * This is best effort
+	 * We are inevitably racing with the hardware.
+	 */
 	if (test_bit(WDM_DISCONNECTING, &desc->flags))
 		return -ENODEV;
 	if (!rv)
 		return -EIO;
 
 	rv = desc->werr;
-	if (rv < 0)
-		dev_err(&desc->intf->dev, "Error in flush path: %d\n",
-			rv);
 
 	return usb_translate_errors(rv);
 }
Signature · 1 line
2.16.4
#4 lore raw

RFC 4/7 CDC-WDM: fix race reporting errors in flush

In reply to Oliver Neukum

In case a race was lost and multiple fds used,
an error could be reported multiple times. To fix
this a spinlock must be taken.

---
 drivers/usb/class/cdc-wdm.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/usb/class/cdc-wdm.c +3 -0
index 1ca2c85a977d..5fb855404403 100644
@@ -613,7 +613,10 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 	if (!rv)
 		return -EIO;
 
+	spin_lock_irq(&desc->iuspin);
 	rv = desc->werr;
+	desc->werr = 0;
+	spin_unlock_irq(&desc->iuspin);
 
 	return usb_translate_errors(rv);
 }
Signature · 1 line
2.16.4
#5 lore raw

RFC 5/7 CDC-WDM: correct error reporting in write()

In reply to Oliver Neukum

In case our wait was interrupted by a disconnect, we should report that.

---
 drivers/usb/class/cdc-wdm.c | 3 +++
 1 file changed, 3 insertions(+)
drivers/usb/class/cdc-wdm.c +3 -0
index 5fb855404403..242f83118208 100644
@@ -396,6 +396,9 @@ static ssize_t wdm_write
 	if (test_bit(WDM_RESETTING, &desc->flags))
 		r = -EIO;
 
+	if (test_bit(WDM_DISCONNECTING, &desc->flags))
+		r = -ENODEV;
+
 	if (r < 0) {
 		rv = r;
 		goto out_free_mem_pm;
Signature · 1 line
2.16.4
#6 lore raw

RFC 6/7 CDC-WDM: implement fsync

In reply to Oliver Neukum

Some users want to be very sure that data has gone out to the device. This needs fsync.

---
 drivers/usb/class/cdc-wdm.c | 28 ++++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)
drivers/usb/class/cdc-wdm.c +28 -0
index 242f83118208..6ea03c12380c 100644
@@ -590,6 +590,33 @@ static ssize_t wdm_read
 	return rv;
 }
 
+/*
+ * The difference to flush is that we wait forever. If you don't like
+ * that behavior, you need to send a signal.
+ */
+
+static int wdm_fsync(struct file *file, loff_t start, loff_t end, int datasync)
+{
+	struct wdm_device *desc = file->private_data;
+	int rv;
+
+	rv = wait_event_interruptible(desc->wait,
+			!test_bit(WDM_IN_USE, &desc->flags) ||
+			test_bit(WDM_DISCONNECTING, &desc->flags));
+
+	if (test_bit(WDM_DISCONNECTING, &desc->flags))
+		return -ENODEV;
+	if (rv < 0)
+		return -EINTR;
+
+	spin_lock_irq(&desc->iuspin);
+	rv = desc->werr;
+	desc->werr = 0;
+	spin_unlock_irq(&desc->iuspin);
+
+	return usb_translate_errors(rv);
+}
+
 static int wdm_flush(struct file *file, fl_owner_t id)
 {
 	struct wdm_device *desc = file->private_data;
@@ -746,6 +773,7 @@ static const struct file_operations wdm_fops = {
 	.owner =	THIS_MODULE,
 	.read =		wdm_read,
 	.write =	wdm_write,
+	.fsync =	wdm_fsync,
 	.open =		wdm_open,
 	.flush =	wdm_flush,
 	.release =	wdm_release,
Signature · 1 line
2.16.4
#7 lore raw

RFC 7/7 CDC-WDM: making flush() interruptible

In reply to Oliver Neukum

There is no need for flush() to be uninterruptible. close(2) is allowed to return -EINTR.

30 seconds is quite long a time to sleep in an uninterruptible state. Change it to an interruptible sleep.

---
 drivers/usb/class/cdc-wdm.c | 4 +++-
 1 file changed, 3 insertions(+), 1 deletion(-)
drivers/usb/class/cdc-wdm.c +3 -1
index 6ea03c12380c..b9cca1cb5058 100644
@@ -622,7 +622,7 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 	struct wdm_device *desc = file->private_data;
 	int rv;
 
-	rv = wait_event_timeout(desc->wait,
+	rv = wait_event_interruptible_timeout(desc->wait,
 			/*
 			 * needs both flags. We cannot do with one
 			 * because resetting it would cause a race
@@ -642,6 +642,8 @@ static int wdm_flush(struct file *file, fl_owner_t id)
 		return -ENODEV;
 	if (!rv)
 		return -EIO;
+	if (rv < 0)
+		return -EINTR;
 
 	spin_lock_irq(&desc->iuspin);
 	rv = desc->werr;
Signature · 1 line
2.16.4
#8 lore raw

Re: [RFC 7/7] CDC-WDM: making flush() interruptible

Quoted 5 lines · Oliver Neukum

There is no need for flush() to be uninterruptible. close(2) is allowed to return -EINTR.

30 seconds is quite long a time to sleep in an uninterruptible state. Change it to an interruptible sleep.

Doesn't this conflict with

Making the wait for IO interruptible would not solve the issue. While it would avoid a hang, it would not allow any progress and we would end up with an unclosable fd.

in [RFC 2/7] ? I suggested killable version, for giving up upon SIGKILL implies automatically closing fds by terminating that userspace process.

#9 lore raw

Re: [RFC 7/7] CDC-WDM: making flush() interruptible

Am Dienstag, den 22.09.2020, 20:38 +0900 schrieb Tetsuo Handa:

Quoted 7 lines · Oliver Neukum
Quoted 5 lines

There is no need for flush() to be uninterruptible. close(2) is allowed to return -EINTR.

30 seconds is quite long a time to sleep in an uninterruptible state. Change it to an interruptible sleep.

Quoted 8 lines

Doesn't this conflict with

Making the wait for IO interruptible would not solve the issue. While it would avoid a hang, it would not allow any progress and we would end up with an unclosable fd.

in [RFC 2/7] ? I suggested killable version, for giving up upon SIGKILL implies automatically closing fds by terminating that userspace process.

No. That we state in an earlier patch that an issue needs a timeout instead of an interruptible sleep, does not mean that another issue could not be solved by using an interruptible sleep.

Regards
Oliver

#10 lore raw
In reply to Oliver Neukum
Quoted 12 lines · Oliver Neukum

Stress testing has shown that CDC-WDM has some issues with hangs and error reporting

1. wakeups are not correctly handled in multhreaded environments
2. unresponsive hardware is not handled
3. errors are not correctly reported. This needs flush() to be implemented.

This version makes wdm_flush() use interruptible sleep.

For easier review all squashed together:

I have like 3 or 4 different "RFC" series here from you for this driver, which one is the "newest"?

And can you send a series that isn't RFC so that I can know you feel it is good enough to be merged?

thanks,

greg k-h

#11 lore raw
Quoted 9 lines · Greg KH
Quoted 8 lines · Oliver Neukum

Stress testing has shown that CDC-WDM has some issues with hangs and error reporting

1. wakeups are not correctly handled in multhreaded environments
2. unresponsive hardware is not handled
3. errors are not correctly reported. This needs flush() to be implemented.

Quoted 7 lines
Quoted 4 lines

This version makes wdm_flush() use interruptible sleep.

For easier review all squashed together:

I have like 3 or 4 different "RFC" series here from you for this driver, which one is the "newest"?

https://lkml.kernel.org/r/[email protected]

is the newest series from Oliver. But

https://lkml.kernel.org/r/[email protected]

is the squashed version with updated comments and deduplicated code.

Quoted 3 lines

And can you send a series that isn't RFC so that I can know you feel it is good enough to be merged?

Do you want this fix as a series of patches (the former link)?
Since I think that the changeset should be atomically applied, I posted the latter link.

#12 lore raw
Quoted 24 lines · Tetsuo Handa
Quoted 9 lines · Greg KH
Quoted 8 lines · Oliver Neukum

Stress testing has shown that CDC-WDM has some issues with hangs and error reporting

1. wakeups are not correctly handled in multhreaded environments
2. unresponsive hardware is not handled
3. errors are not correctly reported. This needs flush() to be implemented.

Quoted 7 lines
Quoted 4 lines

This version makes wdm_flush() use interruptible sleep.

For easier review all squashed together:

I have like 3 or 4 different "RFC" series here from you for this driver, which one is the "newest"?

https://lkml.kernel.org/r/[email protected]

is the newest series from Oliver. But

https://lkml.kernel.org/r/[email protected]

Quoted 8 lines

is the squashed version with updated comments and deduplicated code.

Quoted 3 lines

And can you send a series that isn't RFC so that I can know you feel it is good enough to be merged?

Do you want this fix as a series of patches (the former link)?
Since I think that the changeset should be atomically applied, I posted the latter link.

A single patch would be good to send to me again, burried at the end of a long thread is hard to dig out.

Also with proper authorship is needed, did Oliver write this, or did you?

There is the co-developed-by: tag, which looks like it might be relevant here, can you do that?

thanks,

greg k-h

#13 lore raw
Quoted 8 lines · Greg KH
Quoted 2 lines

Do you want this fix as a series of patches (the former link)?
Since I think that the changeset should be atomically applied, I posted the latter link.

A single patch would be good to send to me again, burried at the end of a long thread is hard to dig out.

Also with proper authorship is needed, did Oliver write this, or did you?

Oliver wrote the majority part. I just squashed the series and updated comments and deduplicated the code. Thus, I think main authorship should be given to Oliver.

Quoted 3 lines

There is the co-developed-by: tag, which looks like it might be relevant here, can you do that?

If you prefer the co-developed-by: tag, you can add:

Oliver, any comments?

#14 lore raw
Quoted 14 lines · Tetsuo Handa
Quoted 8 lines · Greg KH
Quoted 2 lines

Do you want this fix as a series of patches (the former link)?
Since I think that the changeset should be atomically applied, I posted the latter link.

A single patch would be good to send to me again, burried at the end of a long thread is hard to dig out.

Also with proper authorship is needed, did Oliver write this, or did you?

Oliver wrote the majority part. I just squashed the series and updated comments and deduplicated the code. Thus, I think main authorship should be given to Oliver.

Quoted 1 line
Quoted 6 lines
Quoted 2 lines

There is the co-developed-by: tag, which looks like it might be relevant here, can you do that?

If you prefer the co-developed-by: tag, you can add:

That seems reasonable, then put Oliver's address on the first line with a "From:" line to make this all work correctly when I apply it.

thanks,

greg k-h

#15 lore raw

Am Samstag, den 26.09.2020, 07:40 +0200 schrieb Greg KH:

Quoted 17 lines · Tetsuo Handa
Quoted 16 lines · Greg KH
Quoted 8 lines
Quoted 2 lines

Do you want this fix as a series of patches (the former link)?
Since I think that the changeset should be atomically applied, I posted the latter link.

A single patch would be good to send to me again, burried at the end of a long thread is hard to dig out.

Also with proper authorship is needed, did Oliver write this, or did you?

Oliver wrote the majority part. I just squashed the series and updated comments and deduplicated the code. Thus, I think main authorship should be given to Oliver.

Quoted 2 lines

There is the co-developed-by: tag, which looks like it might be relevant here, can you do that?

Quoted 6 lines
Quoted 3 lines

If you prefer the co-developed-by: tag, you can add:

That seems reasonable, then put Oliver's address on the first line with a "From:" line to make this all work correctly when I apply it.

Hi,

I did sent out the series with a PATCH label last week. Should I resend?

Regards
Oliver

#16 lore raw
Quoted 8 lines · Oliver Neukum
Quoted 2 lines

That seems reasonable, then put Oliver's address on the first line with a "From:" line to make this all work correctly when I apply it.

Hi,

I did sent out the series with a PATCH label last week. Should I resend?

No need to resend. I already sent a squashed version as
https://lkml.kernel.org/r/[email protected] .