35022: hid-magicmouse broken

35022: hid-magicmouse broken

[Chase Douglas]

Hi Jiri,

I've received a few emails about the magicmouse driver being broken on
2.6.39:

https://bugzilla.kernel.org/show_bug.cgi?id=35022

It seems hid_output_raw_report() is returning a different (incorrect?)
value than in the past. Do you know what might be going on?

Thanks!

-- Chase

Re: 35022: hid-magicmouse broken

[Jiri Kosina]

On Tue, 17 May 2011, Chase Douglas wrote:

> Hi Jiri,
> 
> I've received a few emails about the magicmouse driver being broken on
> 2.6.39:
> 
> https://bugzilla.kernel.org/show_bug.cgi?id=35022
> 
> It seems hid_output_raw_report() is returning a different (incorrect?)
> value than in the past. Do you know what might be going on?

So we are getting EIO from the transport-level _raw callback.

Hmm, commit 0825411ade seems like a suspect here. Might be possible that 
magicmouse doesn't send ACK back, right?

Could you please try reverting that commit and re-test?

Adding Alan to CC.

-- 
Jiri Kosina
SUSE Labs

Re: 35022: hid-magicmouse broken

[Chase Douglas]

On 05/17/2011 08:56 AM, Jiri Kosina wrote:
> On Tue, 17 May 2011, Chase Douglas wrote:
> 
>> Hi Jiri,
>>
>> I've received a few emails about the magicmouse driver being broken on
>> 2.6.39:
>>
>> https://bugzilla.kernel.org/show_bug.cgi?id=35022
>>
>> It seems hid_output_raw_report() is returning a different (incorrect?)
>> value than in the past. Do you know what might be going on?
> 
> So we are getting EIO from the transport-level _raw callback.
> 
> Hmm, commit 0825411ade seems like a suspect here. Might be possible that 
> magicmouse doesn't send ACK back, right?
> 
> Could you please try reverting that commit and re-test?

Yes, reverting that commit makes it work.

I'm just speculating here, based on commit message names and what you've
said, that the magicmouse is not protocol compliant because it is not
sending an ACK back? What do you think we should do in the driver?
Should we ignore the return, or should we look specifically for EIO?
(neither sounds very good to me, so I'm hoping you have a better
solution :).

Thanks!

-- Chase

Re: 35022: hid-magicmouse broken

[Jiri Kosina]

On Tue, 17 May 2011, Chase Douglas wrote:

> >> It seems hid_output_raw_report() is returning a different (incorrect?)
> >> value than in the past. Do you know what might be going on?
> > 
> > So we are getting EIO from the transport-level _raw callback.
> > 
> > Hmm, commit 0825411ade seems like a suspect here. Might be possible that 
> > magicmouse doesn't send ACK back, right?
> > 
> > Could you please try reverting that commit and re-test?
> 
> Yes, reverting that commit makes it work.
> 
> I'm just speculating here, based on commit message names and what you've
> said, that the magicmouse is not protocol compliant because it is not
> sending an ACK back? 

Yes, I believe that's what is happening. Could you please report what is 
the dmesg output with the patch below, just to make sure that we 
understand the situation precisely?

> What do you think we should do in the driver? Should we ignore the 
> return, or should we look specifically for EIO? (neither sounds very 
> good to me, so I'm hoping you have a better solution :).

Well if the device indeed doesn't send the ACK and it should (I will have 
to check the specs first), we'll have to put a quirk into the driver. 
Otherwise if the ACK is not mandatory, we'll have to revert that commit 
(or at least make it non-fatal failure).

But I'll have to check the specs first.


 net/bluetooth/hidp/core.c |    5 ++++-
 1 files changed, 4 insertions(+), 1 deletions(-)

diff --git a/net/bluetooth/hidp/core.c b/net/bluetooth/hidp/core.c
index 5ec1297..21025ed 100644
--- a/net/bluetooth/hidp/core.c
+++ b/net/bluetooth/hidp/core.c
@@ -403,6 +403,7 @@ static int hidp_output_raw_report(struct hid_device *hid, unsigned char *data, s
 	struct hidp_session *session = hid->driver_data;
 	int ret;
 
+	printk(KERN_DEBUG "hidp_output_raw_report, report_type: %d\n", report_type);
 	switch (report_type) {
 	case HID_FEATURE_REPORT:
 		report_type = HIDP_TRANS_SET_REPORT | HIDP_DATA_RTYPE_FEATURE;
@@ -434,6 +435,7 @@ static int hidp_output_raw_report(struct hid_device *hid, unsigned char *data, s
 			10*HZ);
 		if (res == 0) {
 			/* timeout */
+			printk(KERN_DEBUG "hidp: returning -EIO because of timeout\n");
 			ret = -EIO;
 			goto err;
 		}
@@ -445,6 +447,7 @@ static int hidp_output_raw_report(struct hid_device *hid, unsigned char *data, s
 	}
 
 	if (!session->output_report_success) {
+		printk(KERN_DEBUG "hidp: returnign -EIO because of !output_report_success\n");
 		ret = -EIO;
 		goto err;
 	}
@@ -480,7 +483,7 @@ static inline void hidp_del_timer(struct hidp_session *session)
 static void hidp_process_handshake(struct hidp_session *session,
 					unsigned char param)
 {
-	BT_DBG("session %p param 0x%02x", session, param);
+	printk(KERN_DEBUG "session %p param 0x%02x", session, param);
 	session->output_report_success = 0; /* default condition */
 
 	switch (param) {

-- 
Jiri Kosina
SUSE Labs

Re: 35022: hid-magicmouse broken

[Chase Douglas]

On 05/19/2011 07:59 AM, Jiri Kosina wrote:
> On Tue, 17 May 2011, Chase Douglas wrote:
> 
>>>> It seems hid_output_raw_report() is returning a different (incorrect?)
>>>> value than in the past. Do you know what might be going on?
>>>
>>> So we are getting EIO from the transport-level _raw callback.
>>>
>>> Hmm, commit 0825411ade seems like a suspect here. Might be possible that 
>>> magicmouse doesn't send ACK back, right?
>>>
>>> Could you please try reverting that commit and re-test?
>>
>> Yes, reverting that commit makes it work.
>>
>> I'm just speculating here, based on commit message names and what you've
>> said, that the magicmouse is not protocol compliant because it is not
>> sending an ACK back? 
> 
> Yes, I believe that's what is happening. Could you please report what is 
> the dmesg output with the patch below, just to make sure that we 
> understand the situation precisely?

Here's the output in dmesg:

[  195.491735] Bluetooth: HIDP (Human Interface Emulation) ver 1.2
[  222.693947] input: cndougla’s trackpad as
/devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.3/1-1.3.3/1-1.3.3:1.0/bluetooth/hci0/hci0:12/input16
[  222.694119] magicmouse 0005:05AC:030E.0005: input,hidraw3: BLUETOOTH
HID v1.60 Mouse [cndougla’s trackpad] on 50:63:13:90:AF:A4
[  222.694128] hidp_output_raw_report, report_type: 2
[  222.808111] session ffff88012d9b1200 param 0x02
[  222.808198] hidp: returnign -EIO because of !output_report_success
[  222.808209] magicmouse 0005:05AC:030E.0005: unable to request touch
data (-5)
[  222.809358] session ffff88012d9b1200 param 0x02
[  222.810606] session ffff88012d9b1200 param 0x02
[  222.813113] session ffff88012d9b1200 param 0x00
[  223.045363] magicmouse: probe of 0005:05AC:030E.0005 failed with error -5

>> What do you think we should do in the driver? Should we ignore the 
>> return, or should we look specifically for EIO? (neither sounds very 
>> good to me, so I'm hoping you have a better solution :).
> 
> Well if the device indeed doesn't send the ACK and it should (I will have 
> to check the specs first), we'll have to put a quirk into the driver. 
> Otherwise if the ACK is not mandatory, we'll have to revert that commit 
> (or at least make it non-fatal failure).
> 
> But I'll have to check the specs first.

Sounds good to me :).

Thanks!

-- Chase
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: 35022: hid-magicmouse broken

[Jiri Kosina]

On Thu, 19 May 2011, Chase Douglas wrote:

> Here's the output in dmesg:
> 
> [  195.491735] Bluetooth: HIDP (Human Interface Emulation) ver 1.2
> [  222.693947] input: cndougla’s trackpad as
> /devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.3/1-1.3.3/1-1.3.3:1.0/bluetooth/hci0/hci0:12/input16
> [  222.694119] magicmouse 0005:05AC:030E.0005: input,hidraw3: BLUETOOTH
> HID v1.60 Mouse [cndougla’s trackpad] on 50:63:13:90:AF:A4
> [  222.694128] hidp_output_raw_report, report_type: 2
> [  222.808111] session ffff88012d9b1200 param 0x02
> [  222.808198] hidp: returnign -EIO because of !output_report_success
> [  222.808209] magicmouse 0005:05AC:030E.0005: unable to request touch
> data (-5)
> [  222.809358] session ffff88012d9b1200 param 0x02
> [  222.810606] session ffff88012d9b1200 param 0x02
> [  222.813113] session ffff88012d9b1200 param 0x00
> [  223.045363] magicmouse: probe of 0005:05AC:030E.0005 failed with error -5

Ok, so what is happening here is that magicmouse_probe() sends the { 0xd7, 
0x01 } feature report to the device, and it responds with 
HIDP_HSHK_ERR_INVALID_REPORT_ID.

That's why the _raw callback (correctly) returns error.

So the question to the driver authors is -- what is the point behind the { 
0xd7, 0x01 } report? Clearly the device doesn't like it during probe 
because of invalid report ID.
And it never did, we just silently ignored the error with the older 
kernels.

Michael?

-- 
Jiri Kosina
SUSE Labs
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: 35022: hid-magicmouse broken

[Chase Douglas]

On 05/19/2011 10:54 AM, Jiri Kosina wrote:
> On Thu, 19 May 2011, Chase Douglas wrote:
> 
>> Here's the output in dmesg:
>>
>> [  195.491735] Bluetooth: HIDP (Human Interface Emulation) ver 1.2
>> [  222.693947] input: cndougla’s trackpad as
>> /devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.3/1-1.3.3/1-1.3.3:1.0/bluetooth/hci0/hci0:12/input16
>> [  222.694119] magicmouse 0005:05AC:030E.0005: input,hidraw3: BLUETOOTH
>> HID v1.60 Mouse [cndougla’s trackpad] on 50:63:13:90:AF:A4
>> [  222.694128] hidp_output_raw_report, report_type: 2
>> [  222.808111] session ffff88012d9b1200 param 0x02
>> [  222.808198] hidp: returnign -EIO because of !output_report_success
>> [  222.808209] magicmouse 0005:05AC:030E.0005: unable to request touch
>> data (-5)
>> [  222.809358] session ffff88012d9b1200 param 0x02
>> [  222.810606] session ffff88012d9b1200 param 0x02
>> [  222.813113] session ffff88012d9b1200 param 0x00
>> [  223.045363] magicmouse: probe of 0005:05AC:030E.0005 failed with error -5
> 
> Ok, so what is happening here is that magicmouse_probe() sends the { 0xd7, 
> 0x01 } feature report to the device, and it responds with 
> HIDP_HSHK_ERR_INVALID_REPORT_ID.
> 
> That's why the _raw callback (correctly) returns error.
> 
> So the question to the driver authors is -- what is the point behind the { 
> 0xd7, 0x01 } report? Clearly the device doesn't like it during probe 
> because of invalid report ID.
> And it never did, we just silently ignored the error with the older 
> kernels.

The feature report is what flips the devices (magic mouse, magic
trackpad) into multitouch mode. Without the report, the mouse will not
emit the location of any touches on its surface, and the trackpad will
only emit single touch data.

What do you think we should do? We can't get rid of the report. Should
we silently ignore the error?

Thanks
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: 35022: hid-magicmouse broken

[Jiri Kosina]

On Thu, 19 May 2011, Chase Douglas wrote:

> >> [  195.491735] Bluetooth: HIDP (Human Interface Emulation) ver 1.2
> >> [  222.693947] input: cndougla’s trackpad as
> >> /devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.3/1-1.3.3/1-1.3.3:1.0/bluetooth/hci0/hci0:12/input16
> >> [  222.694119] magicmouse 0005:05AC:030E.0005: input,hidraw3: BLUETOOTH
> >> HID v1.60 Mouse [cndougla’s trackpad] on 50:63:13:90:AF:A4
> >> [  222.694128] hidp_output_raw_report, report_type: 2
> >> [  222.808111] session ffff88012d9b1200 param 0x02
> >> [  222.808198] hidp: returnign -EIO because of !output_report_success
> >> [  222.808209] magicmouse 0005:05AC:030E.0005: unable to request touch
> >> data (-5)
> >> [  222.809358] session ffff88012d9b1200 param 0x02
> >> [  222.810606] session ffff88012d9b1200 param 0x02
> >> [  222.813113] session ffff88012d9b1200 param 0x00
> >> [  223.045363] magicmouse: probe of 0005:05AC:030E.0005 failed with error -5
> > 
> > Ok, so what is happening here is that magicmouse_probe() sends the { 0xd7, 
> > 0x01 } feature report to the device, and it responds with 
> > HIDP_HSHK_ERR_INVALID_REPORT_ID.
> > 
> > That's why the _raw callback (correctly) returns error.
> > 
> > So the question to the driver authors is -- what is the point behind the { 
> > 0xd7, 0x01 } report? Clearly the device doesn't like it during probe 
> > because of invalid report ID.
> > And it never did, we just silently ignored the error with the older 
> > kernels.
> 
> The feature report is what flips the devices (magic mouse, magic
> trackpad) into multitouch mode. Without the report, the mouse will not
> emit the location of any touches on its surface, and the trackpad will
> only emit single touch data.
> 
> What do you think we should do? We can't get rid of the report. Should
> we silently ignore the error?

Hmm, I am afraid that it's the only option (apart from thowing the whole 
'waiting for ack' infrastructure away, which would be sad, as it's a good 
thing in principle).

So how about the patch below? :/ Could you please verify that it makes 
hid-magicmouse work again? Thanks.



From: Jiri Kosina <jkosina@suse.cz>
Subject: [PATCH] HID: magicmouse: ignore 'ivalid report id' while switching modes

The device reponds with 'invalid report id' when feature report switching it
into multitouch mode is sent to it.

This has been silently ignored before 0825411ade ("HID: bt: Wait for ACK
on Sent Reports"), but since this commit, it propagates -EIO from the _raw
callback .

So let the driver ignore -EIO as response to 0xd7,0x01 report, as that's
how the device reacts in normal mode.

Sad, but following reality.

This fixes https://bugzilla.kernel.org/show_bug.cgi?id=35022

Signed-off-by: Jiri Kosina <jkosina@suse.cz>
---
 drivers/hid/hid-magicmouse.c |   10 +++++++++-
 1 files changed, 9 insertions(+), 1 deletions(-)

diff --git a/drivers/hid/hid-magicmouse.c b/drivers/hid/hid-magicmouse.c
index 0ec91c1..a5eda4c 100644
--- a/drivers/hid/hid-magicmouse.c
+++ b/drivers/hid/hid-magicmouse.c
@@ -501,9 +501,17 @@ static int magicmouse_probe(struct hid_device *hdev,
 	}
 	report->size = 6;
 
+	/*
+	 * The device reponds with 'invalid report id' when feature
+	 * report switching it into multitouch mode is sent to it.
+	 *
+	 * This results in -EIO from the _raw low-level transport callback,
+	 * but there seems to be no other way of switching the mode.
+	 * Thus the super-ugly hacky success check below.
+	 */
 	ret = hdev->hid_output_raw_report(hdev, feature, sizeof(feature),
 			HID_FEATURE_REPORT);
-	if (ret != sizeof(feature)) {
+	if (ret != -EIO) {
 		hid_err(hdev, "unable to request touch data (%d)\n", ret);
 		goto err_stop_hw;
 	}
-- 
1.7.3.1

--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: 35022: hid-magicmouse broken

[Chase Douglas]

On 05/19/2011 12:03 PM, Jiri Kosina wrote:
> On Thu, 19 May 2011, Chase Douglas wrote:
> 
>>>> [  195.491735] Bluetooth: HIDP (Human Interface Emulation) ver 1.2
>>>> [  222.693947] input: cndougla’s trackpad as
>>>> /devices/pci0000:00/0000:00:1a.0/usb1/1-1/1-1.3/1-1.3.3/1-1.3.3:1.0/bluetooth/hci0/hci0:12/input16
>>>> [  222.694119] magicmouse 0005:05AC:030E.0005: input,hidraw3: BLUETOOTH
>>>> HID v1.60 Mouse [cndougla’s trackpad] on 50:63:13:90:AF:A4
>>>> [  222.694128] hidp_output_raw_report, report_type: 2
>>>> [  222.808111] session ffff88012d9b1200 param 0x02
>>>> [  222.808198] hidp: returnign -EIO because of !output_report_success
>>>> [  222.808209] magicmouse 0005:05AC:030E.0005: unable to request touch
>>>> data (-5)
>>>> [  222.809358] session ffff88012d9b1200 param 0x02
>>>> [  222.810606] session ffff88012d9b1200 param 0x02
>>>> [  222.813113] session ffff88012d9b1200 param 0x00
>>>> [  223.045363] magicmouse: probe of 0005:05AC:030E.0005 failed with error -5
>>>
>>> Ok, so what is happening here is that magicmouse_probe() sends the { 0xd7, 
>>> 0x01 } feature report to the device, and it responds with 
>>> HIDP_HSHK_ERR_INVALID_REPORT_ID.
>>>
>>> That's why the _raw callback (correctly) returns error.
>>>
>>> So the question to the driver authors is -- what is the point behind the { 
>>> 0xd7, 0x01 } report? Clearly the device doesn't like it during probe 
>>> because of invalid report ID.
>>> And it never did, we just silently ignored the error with the older 
>>> kernels.
>>
>> The feature report is what flips the devices (magic mouse, magic
>> trackpad) into multitouch mode. Without the report, the mouse will not
>> emit the location of any touches on its surface, and the trackpad will
>> only emit single touch data.
>>
>> What do you think we should do? We can't get rid of the report. Should
>> we silently ignore the error?
> 
> Hmm, I am afraid that it's the only option (apart from thowing the whole 
> 'waiting for ack' infrastructure away, which would be sad, as it's a good 
> thing in principle).
> 
> So how about the patch below? :/ Could you please verify that it makes 
> hid-magicmouse work again? Thanks.
> 
> 
> 
> From: Jiri Kosina <jkosina@suse.cz>
> Subject: [PATCH] HID: magicmouse: ignore 'ivalid report id' while switching modes
> 
> The device reponds with 'invalid report id' when feature report switching it
> into multitouch mode is sent to it.
> 
> This has been silently ignored before 0825411ade ("HID: bt: Wait for ACK
> on Sent Reports"), but since this commit, it propagates -EIO from the _raw
> callback .
> 
> So let the driver ignore -EIO as response to 0xd7,0x01 report, as that's
> how the device reacts in normal mode.
> 
> Sad, but following reality.
> 
> This fixes https://bugzilla.kernel.org/show_bug.cgi?id=35022
> 
> Signed-off-by: Jiri Kosina <jkosina@suse.cz>
> ---
>  drivers/hid/hid-magicmouse.c |   10 +++++++++-
>  1 files changed, 9 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/hid/hid-magicmouse.c b/drivers/hid/hid-magicmouse.c
> index 0ec91c1..a5eda4c 100644
> --- a/drivers/hid/hid-magicmouse.c
> +++ b/drivers/hid/hid-magicmouse.c
> @@ -501,9 +501,17 @@ static int magicmouse_probe(struct hid_device *hdev,
>  	}
>  	report->size = 6;
>  
> +	/*
> +	 * The device reponds with 'invalid report id' when feature
> +	 * report switching it into multitouch mode is sent to it.
> +	 *
> +	 * This results in -EIO from the _raw low-level transport callback,
> +	 * but there seems to be no other way of switching the mode.
> +	 * Thus the super-ugly hacky success check below.
> +	 */
>  	ret = hdev->hid_output_raw_report(hdev, feature, sizeof(feature),
>  			HID_FEATURE_REPORT);
> -	if (ret != sizeof(feature)) {
> +	if (ret != -EIO) {
>  		hid_err(hdev, "unable to request touch data (%d)\n", ret);
>  		goto err_stop_hw;
>  	}

Seems to work for me :).

Reviewed-by: Chase Douglas <chase.douglas@canonical.com>

Thanks!
--
To unsubscribe from this list: send the line "unsubscribe linux-input" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html

Re: 35022: hid-magicmouse broken

[Jiri Kosina]

On Thu, 19 May 2011, Chase Douglas wrote:

> Seems to work for me :).

I have applied the patch with your Tested-by. Thanks,

-- 
Jiri Kosina
SUSE Labs

Re: 35022: hid-magicmouse broken

[Chase Douglas]

On 05/20/2011 04:28 AM, Jiri Kosina wrote:
> On Thu, 19 May 2011, Chase Douglas wrote:
> 
>> Seems to work for me :).
> 
> I have applied the patch with your Tested-by. Thanks,

Thanks Jiri!

Could you add 'Cc: stable@kernel.org' if it hasn't been merged yet? It
looks like .39 was just released without this patch...

Re: 35022: hid-magicmouse broken

[Jiri Kosina]

On Fri, 20 May 2011, Chase Douglas wrote:

> >> Seems to work for me :).
> > 
> > I have applied the patch with your Tested-by. Thanks,
> 
> Thanks Jiri!
> 
> Could you add 'Cc: stable@kernel.org' if it hasn't been merged yet? It
> looks like .39 was just released without this patch...

I will be sending this patch to stable once it's in Linus tree (I will be 
sending pull request likely today).

-- 
Jiri Kosina
SUSE Labs