On Oct 06 2023, Hans de Goede wrote:

Hi Benjamin,

On 10/6/23 15:46, Benjamin Tissoires wrote:

...

Hi Hans,

On Oct 06 2023, Hans de Goede wrote:

...

hidpp_probe() restarts IO after setting things up, if we get a connect event just before hidpp_probe() stops all IO then hidpp_connect_event() will see IO errors causing it to fail to setup the connected device.

I think I see why you are doing this, but it scares me to be honest.

...

Add a new io_mutex which hidpp_probe() locks while restarting IO and which is also taken by hidpp_connect_event() to avoid these 2 things from racing.

So now we are adding a new mutex to prevent a workqueue to be executed, which is held while there is another semaphore going down/up... It feels error prone to say the least and I'm not sure we are not actually fixing the problem.

My guts tells me that the issue is tackled at the wrong time, and that maybe there is a better solution that doesn't involve a new lock in the middle of 2 other locks being actually held...

Since the lock is only taken into 2 places and 1 of them is holding no locks when taking it (because workqueue) I don't really see how this would be a problem.

Actually introducing a new lock for this, rather then say trying to use the send_mutex makes this much easier to reason about, more on this below.

...

One minor comment in the code.

...

Hopefully this will help with the occasional connect errors leading to e.g. missing battery monitoring.

Cc: stable@vger.kernel.org Signed-off-by: Hans de Goede hdegoede@redhat.com

---

drivers/hid/hid-logitech-hidpp.c | 35 ++++++++++++++++++++++---------- 1 file changed, 24 insertions(+), 11 deletions(-)

diff --git a/drivers/hid/hid-logitech-hidpp.c b/drivers/hid/hid-logitech-hidpp.c index a209d51bd247..33f9cd98485a 100644 --- a/drivers/hid/hid-logitech-hidpp.c +++ b/drivers/hid/hid-logitech-hidpp.c @@ -181,6 +181,7 @@ struct hidpp_scroll_counter { struct hidpp_device { struct hid_device *hid_dev; struct input_dev *input;

• struct mutex io_mutex; struct mutex send_mutex; void *send_receive_buf; char *name; /* will never be NULL and should not be freed */

@@ -4207,36 +4208,39 @@ static void hidpp_connect_event(struct hidpp_device *hidpp) return; }

• /* Avoid probe() restarting IO */
• mutex_lock(&hidpp->io_mutex);

I'd put a `__must_hold(&hidpp->io_mutex);` here, not changing any return path and forcing any caller to `hidpp_connect_event()` (which will eventually only be the work struct) to take the lock.

This should simplify the patch by a lot and also ensure someone doesn't forget the `goto out_unlock`.

Ok, I can add the __must_hold() here and make delayed_Work_cb take the lock, but that would make it impossible to implement patch 2/2 in a clean manner and I do like patch 2/2 since it makes it clear that hidpp_connect_event must only run from the workqueue but I guess we could just add a comment for that instead.

In 2/2, just rename this function to __do_hidpp_connect_event(), and have hidpp_connect_event() being the worker, which takes the lock, and calls __do_hidpp_connect_event().

Either way works for me, with a slight preference for the current version even if it introduces a bunch of gotos.

...

...

• if (hidpp->quirks & HIDPP_QUIRK_CLASS_WTP) { ret = wtp_connect(hdev, connected); if (ret)

• 	return;
  

• 	goto out_unlock;
  

  } else if (hidpp->quirks & HIDPP_QUIRK_CLASS_M560) { ret = m560_send_config_command(hdev, connected); if (ret)

• 	return;
  

• 	goto out_unlock;
  

  } else if (hidpp->quirks & HIDPP_QUIRK_CLASS_K400) { ret = k400_connect(hdev, connected); if (ret)

• 	return;
  

• 	goto out_unlock;
  

  }

if (hidpp->quirks & HIDPP_QUIRK_HIDPP_WHEELS) { ret = hidpp10_wheel_connect(hidpp); if (ret)

• 	return;
  

• 	goto out_unlock;
  

  }

if (hidpp->quirks & HIDPP_QUIRK_HIDPP_EXTRA_MOUSE_BTNS) { ret = hidpp10_extra_mouse_buttons_connect(hidpp); if (ret)

• 	return;
  

• 	goto out_unlock;
  

  }

if (hidpp->quirks & HIDPP_QUIRK_HIDPP_CONSUMER_VENDOR_KEYS) { ret = hidpp10_consumer_keys_connect(hidpp); if (ret)

• 	return;
  

• 	goto out_unlock;
  

  }

/* the device is already connected, we can ask for its name and @@ -4245,7 +4249,7 @@ static void hidpp_connect_event(struct hidpp_device *hidpp) ret = hidpp_root_get_protocol_version(hidpp); if (ret) { hid_err(hdev, "Can not get the protocol version.\n");

• 	return;
  

• 	goto out_unlock;
  

  } }

@@ -4256,7 +4260,7 @@ static void hidpp_connect_event(struct hidpp_device *hidpp) "%s", name); kfree(name); if (!devm_name)

• 		return;
  

• 		goto out_unlock;
  

hidpp->name = devm_name; } @@ -4291,12 +4295,12 @@ static void hidpp_connect_event(struct hidpp_device *hidpp) if (!(hidpp->quirks & HIDPP_QUIRK_DELAYED_INIT) || hidpp->delayed_input) /* if the input nodes are already created, we can stop now */

• return;
  

• goto out_unlock;
  

input = hidpp_allocate_input(hdev); if (!input) { hid_err(hdev, "cannot allocate new input device: %d\n", ret);

• return;
  

• goto out_unlock;
  

  }

hidpp_populate_input(hidpp, input); @@ -4304,10 +4308,12 @@ static void hidpp_connect_event(struct hidpp_device *hidpp) ret = input_register_device(input); if (ret) { input_free_device(input);

• return;
  

• goto out_unlock;
  

  }

hidpp->delayed_input = input; +out_unlock:

• mutex_unlock(&hidpp->io_mutex);

} static DEVICE_ATTR(builtin_power_supply, 0000, NULL, NULL); @@ -4450,6 +4456,7 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id) will_restart = true; INIT_WORK(&hidpp->work, delayed_work_cb);

• mutex_init(&hidpp->io_mutex); mutex_init(&hidpp->send_mutex); init_waitqueue_head(&hidpp->wait);

@@ -4519,6 +4526,9 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id) flush_work(&hidpp->work); if (will_restart) {

• /* Avoid hidpp_connect_event() running while restarting */
  

• mutex_lock(&hidpp->io_mutex);
  

• /* Reset the HID node state */ hid_device_io_stop(hdev);

That's the part that makes me raise an eyebrow. Because we lock, then release the semaphore to get it back later. Can this induce a dead lock?

Can't we solve that same scenario without a mutex, but forcing either the workqueue to not run or to be finished at this point?

I'm not sure what you are worried about after the mutex_lock the line above we are 100% guaranteed that hidpp_connect_event() is not running and since it is not running it will also not be holding any other locks, so it can not cause any problems.

Agree, but my point is that you are not entirely solving the issue: if now, between hid_device_io_stop() and hid_hw_close() we receive a connect notification from the device, hid_input_report() will return -EBUSY, and we will lose it (it will not be stacked in the workqueue).

I was thinking at adding a flush_work(&hidpp->work) here, instead of the mutex solution, but yours ensures that any connect event already started will be handled properly, which is a plus.

Still if between the mutex lock here we receive a connect event from the device, we still get -EBUSY at the hid-core layer, and so we will lose it. Maybe that's OK because we might re-ask for the device later (I don't remember exactly the code), but my point is that because we add a mutex doesn't mean we will solve all multi-thread problems. So finding a non-mutex solution sometimes is better :)

And the fact that we need to think through every preemption case often means that there is something wrong *elsewhere*.

The other way around if hidpp_connect_event() is running the mutex_lock() above ensures that it will finishes and release any locks before we get here.

There is no way that both code paths can run at the same time with the new lock. And there thus also is no way that they can cause any new not already held locks to be taken while the other side is running.

I actually introduced the new lock because in my mind introducing the new lock allows to easily reason about the impact on other locking (which is none).

I hope this helps explain. As for the making hidpp_connect_event()'s caller take the lock thing, let me know how you want to resolve that. Either way works for me and I guess the less intrusive version of making the caller take the lock is easier to backport so we should probably go that route.

Again, for that particular point, making the function a private one that doesn't have to cope with the mutex will be simpler in this patch and in the future.

Cheers, Benjamin

Regards,

Hans

...

...

hid_hw_close(hdev);

@@ -4529,6 +4539,7 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id) /* Now export the actual inputs and hidraw nodes to the world */ ret = hid_hw_start(hdev, connect_mask);

• mutex_unlock(&hidpp->io_mutex);
  

  if (ret) { hid_err(hdev, "%s:hid_hw_start returned error\n", __func__); goto hid_hw_start_fail;

@@ -4553,6 +4564,7 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id) sysfs_remove_group(&hdev->dev.kobj, &ps_attribute_group); cancel_work_sync(&hidpp->work); mutex_destroy(&hidpp->send_mutex);

• mutex_destroy(&hidpp->io_mutex); return ret;

} @@ -4568,6 +4580,7 @@ static void hidpp_remove(struct hid_device *hdev) hid_hw_stop(hdev); cancel_work_sync(&hidpp->work); mutex_destroy(&hidpp->send_mutex);

• mutex_destroy(&hidpp->io_mutex);

}

#define LDJ_DEVICE(product) \

2.41.0

Cheers, Benjamin

Hi Benjamin,

On 10/6/23 19:24, Hans de Goede wrote:

Hi,

On 10/6/23 18:28, Benjamin Tissoires wrote:

...

On Oct 06 2023, Hans de Goede wrote:

<snip>

...

...

...

...

@@ -4207,36 +4208,39 @@ static void hidpp_connect_event(struct hidpp_device *hidpp) return; }

• /* Avoid probe() restarting IO */
• mutex_lock(&hidpp->io_mutex);

I'd put a `__must_hold(&hidpp->io_mutex);` here, not changing any return path and forcing any caller to `hidpp_connect_event()` (which will eventually only be the work struct) to take the lock.

This should simplify the patch by a lot and also ensure someone doesn't forget the `goto out_unlock`.

Ok, I can add the __must_hold() here and make delayed_Work_cb take the lock, but that would make it impossible to implement patch 2/2 in a clean manner and I do like patch 2/2 since it makes it clear that hidpp_connect_event must only run from the workqueue but I guess we could just add a comment for that instead.

In 2/2, just rename this function to __do_hidpp_connect_event(), and have hidpp_connect_event() being the worker, which takes the lock, and calls __do_hidpp_connect_event().

Ok, will do for v2.

<snip>

...

...

...

...

@@ -4519,6 +4526,9 @@ static int hidpp_probe(struct hid_device *hdev, const struct hid_device_id *id) flush_work(&hidpp->work); if (will_restart) {

• /* Avoid hidpp_connect_event() running while restarting */
  

• mutex_lock(&hidpp->io_mutex);
  

• /* Reset the HID node state */ hid_device_io_stop(hdev);

That's the part that makes me raise an eyebrow. Because we lock, then release the semaphore to get it back later. Can this induce a dead lock?

Can't we solve that same scenario without a mutex, but forcing either the workqueue to not run or to be finished at this point?

I'm not sure what you are worried about after the mutex_lock the line above we are 100% guaranteed that hidpp_connect_event() is not running and since it is not running it will also not be holding any other locks, so it can not cause any problems.

Agree, but my point is that you are not entirely solving the issue: if now, between hid_device_io_stop() and hid_hw_close() we receive a connect notification from the device, hid_input_report() will return -EBUSY, and we will lose it (it will not be stacked in the workqueue).

I was thinking at adding a flush_work(&hidpp->work) here, instead of the mutex solution, but yours ensures that any connect event already started will be handled properly, which is a plus.

Still if between the mutex lock here we receive a connect event from the device, we still get -EBUSY at the hid-core layer, and so we will lose it. Maybe that's OK because we might re-ask for the device later (I don't remember exactly the code), but my point is that because we add a mutex doesn't mean we will solve all multi-thread problems. So finding a non-mutex solution sometimes is better :)

And the fact that we need to think through every preemption case often means that there is something wrong *elsewhere*.

Right, I did consider seeing if we can get rid of the restart altogether, as the whole restarting thing is actually the problem here. AFAICT this is only really necessary in the WTP path since there where we need to know resolution before instantiating the input device.

But atm this is also done for all unifying devices, which seems unnecessary.

Buy we still need the restart anyways for the WTP case, so we need to make it work reliable anyways.

Now that I understand your concern about the missed connect packet, which I agree is a real concern I think I know how to fix this. I'll prepare a new version of this series tomorrow.

Hmm, thinking more about this, if we normally just create the input device right away even for unifying devices and we already always delay the creation for WTP even during the restart:

            if (hidpp->quirks & HIDPP_QUIRK_DELAYED_INIT)
                    connect_mask &= ~HID_CONNECT_HIDINPUT;

Then I wonder why do we even bother to do the restart thing for unifying devices. Do you know what this is based on ?

I guess this might have to do with ensuring the configure commands are send before creating the input + hidraw devices, but if the connect event comes later on then the configuration is already done later on after the input device has already been created ?

So maybe we should indeed just remove the whole restart thing entirely and also always rely on hidpp_connect_event to send the configuration commands, because currently those are send twice if the device is already connnected at probe() time.

The whole restart thing seems to have been added by you in: 91cf9a98ae4127 ("HID: logitech-hidpp: make .probe usbhid capable")

The commit message says that not starting all the HID subdrivers right from the start is necessary because:

"It means that any configuration retrieved after the initial hid_hw_start would not be exposed to user space without races."

I believe this refers to the filling of hdev->name and hdev->uniq by hidpp_unifying_init() (or other similar code paths for non unifying).

It seems that has been broken though by:

498ba20690357 ("HID: logitech-hidpp: Don't restart communication if not necessary")

Which causes input-dev registration to happen before hidpp_unifying_init().

TL;DR:

1. There seems to be a good reason for the restart dance so we need to fix it rather then remove it.

2. 498ba20690357 ("HID: logitech-hidpp: Don't restart communication if not necessary") (from Bastien) re-introduces the race which the restart tries to fix and should be reverted.

Regards,

Hans