[PATCH 0/2] Input: i8042 - add additional TUXEDO devices to i8042 quirk tables

[PATCH 0/2] Input: i8042 - add additional TUXEDO devices to i8042 quirk tables

[Werner Sembach]

Some additional devices needing i8042 quirks for the keyboard to reliably
work after boot and suspend. The first patch are some older devices I found
needing the quirks in our install routine.

The second patch are two devices that I purposly seperated in an own patch,
because it is not clear cut if these should be applied or not:
They need the quirks for the keyboard to not be completly dead by chance
after suspend. However setting the quirk causes the keyboard to lag for the
first few seconds after every boot and sometimes also a little bit after
resume, giving double inputs or missing keystrokes. However this very
quickly goes away again.
Imho the lag issue is less severe then an occasionally completly dead
keyboard. However I'm interested in other opinions. A counter argument is
that the quirks can always be set via boot parameter, but not unset when
they are part of the quirk list.

[PATCH 1/2] Input: i8042 - add additional TUXEDO devices to i8042 quirk tables

[Werner Sembach]

A lot of modern Clevo barebones have touchpad and/or keyboard issues after
suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
have an external PS/2 port so this can safely be set for all of them.

I'm not entirely sure if every device listed really needs all four quirks,
but after testing and production use. No negative effects could be
observed when setting all four.

Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
Cc: stable@vger.kernel.org
---
 drivers/input/serio/i8042-x86ia64io.h | 76 ++++++++++++++++++++++++---
 1 file changed, 68 insertions(+), 8 deletions(-)

diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
index 50b090e77fca..5204a7dd61d4 100644
--- a/drivers/input/serio/i8042-x86ia64io.h
+++ b/drivers/input/serio/i8042-x86ia64io.h
@@ -900,14 +900,6 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
 		},
 		.driver_data = (void *)(SERIO_QUIRK_NOMUX)
 	},
-	{
-		/* Clevo P650RS, 650RP6, Sager NP8152-S, and others */
-		.matches = {
-			DMI_MATCH(DMI_SYS_VENDOR, "Notebook"),
-			DMI_MATCH(DMI_PRODUCT_NAME, "P65xRP"),
-		},
-		.driver_data = (void *)(SERIO_QUIRK_RESET_ALWAYS)
-	},
 	{
 		/* OQO Model 01 */
 		.matches = {
@@ -1162,6 +1154,74 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
 		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
 					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
 	},
+	{
+		/*
+		 * This is only a partial board_name and might be followed by
+		 * another letter or number. DMI_MATCH however does do partial
+		 * matching.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_PRODUCT_NAME, "P65xH"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
+	{
+		/* Clevo P650RS, 650RP6, Sager NP8152-S, and others */
+		.matches = {
+			DMI_MATCH(DMI_PRODUCT_NAME, "P65xRP"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
+	{
+		/*
+		 * This is only a partial board_name and might be followed by
+		 * another letter or number. DMI_MATCH however does do partial
+		 * matching.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_PRODUCT_NAME, "P65_P67H"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
+	{
+		/*
+		 * This is only a partial board_name and might be followed by
+		 * another letter or number. DMI_MATCH however does do partial
+		 * matching.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_PRODUCT_NAME, "P65_67RP"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
+	{
+		/*
+		 * This is only a partial board_name and might be followed by
+		 * another letter or number. DMI_MATCH however does do partial
+		 * matching.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_PRODUCT_NAME, "P65_67RS"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
+	{
+		/*
+		 * This is only a partial board_name and might be followed by
+		 * another letter or number. DMI_MATCH however does do partial
+		 * matching.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_PRODUCT_NAME, "P67xRP"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
 	{
 		.matches = {
 			DMI_MATCH(DMI_BOARD_NAME, "PB50_70DFx,DDx"),
-- 
2.34.1

[PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Werner Sembach]

A lot of modern Clevo barebones have touchpad and/or keyboard issues after
suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
have an external PS/2 port so this can safely be set for all of them.

I'm not entirely sure if every device listed really needs all four quirks,
but after testing and production use. No negative effects could be
observed when setting all four.

Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
boot and sometimes also after resume. However both are required for the
keyboard to not fail completely sometimes after boot or resume.

Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
Cc: stable@vger.kernel.org
---
 drivers/input/serio/i8042-x86ia64io.h | 28 +++++++++++++++++++++++++++
 1 file changed, 28 insertions(+)

diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
index 5204a7dd61d4..9dc0266e5168 100644
--- a/drivers/input/serio/i8042-x86ia64io.h
+++ b/drivers/input/serio/i8042-x86ia64io.h
@@ -1107,6 +1107,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
 		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
 					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
 	},
+	{
+		/*
+		 * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
+		 * the keyboard very laggy for ~5 seconds after boot and
+		 * sometimes also after resume.
+		 * However both are required for the keyboard to not fail
+		 * completely sometimes after boot or resume.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_BOARD_NAME, "N150CU"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
 	{
 		.matches = {
 			DMI_MATCH(DMI_BOARD_NAME, "NH5xAx"),
@@ -1114,6 +1128,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
 		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
 					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
 	},
+	{
+		/*
+		 * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
+		 * the keyboard very laggy for ~5 seconds after boot and
+		 * sometimes also after resume.
+		 * However both are required for the keyboard to not fail
+		 * completely sometimes after boot or resume.
+		 */
+		.matches = {
+			DMI_MATCH(DMI_BOARD_NAME, "NHxxRZQ"),
+		},
+		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
+					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
+	},
 	{
 		.matches = {
 			DMI_MATCH(DMI_BOARD_NAME, "NL5xRU"),
-- 
2.34.1

Re: [PATCH 1/2] Input: i8042 - add additional TUXEDO devices to i8042 quirk tables

[Hans de Goede]

Hi,

On 7/8/22 18:10, Werner Sembach wrote:
> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
> have an external PS/2 port so this can safely be set for all of them.
> 
> I'm not entirely sure if every device listed really needs all four quirks,
> but after testing and production use. No negative effects could be
> observed when setting all four.
> 
> Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
> Cc: stable@vger.kernel.org

Thanks, patch looks good to me:

Reviewed-by: Hans de Goede <hdegoede@redhat.com>

Regards,

Hans



> ---
>  drivers/input/serio/i8042-x86ia64io.h | 76 ++++++++++++++++++++++++---
>  1 file changed, 68 insertions(+), 8 deletions(-)
> 
> diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
> index 50b090e77fca..5204a7dd61d4 100644
> --- a/drivers/input/serio/i8042-x86ia64io.h
> +++ b/drivers/input/serio/i8042-x86ia64io.h
> @@ -900,14 +900,6 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>  		},
>  		.driver_data = (void *)(SERIO_QUIRK_NOMUX)
>  	},
> -	{
> -		/* Clevo P650RS, 650RP6, Sager NP8152-S, and others */
> -		.matches = {
> -			DMI_MATCH(DMI_SYS_VENDOR, "Notebook"),
> -			DMI_MATCH(DMI_PRODUCT_NAME, "P65xRP"),
> -		},
> -		.driver_data = (void *)(SERIO_QUIRK_RESET_ALWAYS)
> -	},
>  	{
>  		/* OQO Model 01 */
>  		.matches = {
> @@ -1162,6 +1154,74 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>  		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>  					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>  	},
> +	{
> +		/*
> +		 * This is only a partial board_name and might be followed by
> +		 * another letter or number. DMI_MATCH however does do partial
> +		 * matching.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_PRODUCT_NAME, "P65xH"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
> +	{
> +		/* Clevo P650RS, 650RP6, Sager NP8152-S, and others */
> +		.matches = {
> +			DMI_MATCH(DMI_PRODUCT_NAME, "P65xRP"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
> +	{
> +		/*
> +		 * This is only a partial board_name and might be followed by
> +		 * another letter or number. DMI_MATCH however does do partial
> +		 * matching.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_PRODUCT_NAME, "P65_P67H"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
> +	{
> +		/*
> +		 * This is only a partial board_name and might be followed by
> +		 * another letter or number. DMI_MATCH however does do partial
> +		 * matching.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_PRODUCT_NAME, "P65_67RP"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
> +	{
> +		/*
> +		 * This is only a partial board_name and might be followed by
> +		 * another letter or number. DMI_MATCH however does do partial
> +		 * matching.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_PRODUCT_NAME, "P65_67RS"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
> +	{
> +		/*
> +		 * This is only a partial board_name and might be followed by
> +		 * another letter or number. DMI_MATCH however does do partial
> +		 * matching.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_PRODUCT_NAME, "P67xRP"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
>  	{
>  		.matches = {
>  			DMI_MATCH(DMI_BOARD_NAME, "PB50_70DFx,DDx"),

Re: [PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Hans de Goede]

Hi,

On 7/8/22 18:10, Werner Sembach wrote:
> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
> have an external PS/2 port so this can safely be set for all of them.
> 
> I'm not entirely sure if every device listed really needs all four quirks,
> but after testing and production use. No negative effects could be
> observed when setting all four.
> 
> Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
> and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
> boot and sometimes also after resume. However both are required for the
> keyboard to not fail completely sometimes after boot or resume.

Hmm, the very laggy bit does not sound good. Have you looked into other
solutions, e.g. what happens if you use just nomux without any of the
other 3 options ?

Regards,

Hans


> 
> Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
> Cc: stable@vger.kernel.org
> ---
>  drivers/input/serio/i8042-x86ia64io.h | 28 +++++++++++++++++++++++++++
>  1 file changed, 28 insertions(+)
> 
> diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
> index 5204a7dd61d4..9dc0266e5168 100644
> --- a/drivers/input/serio/i8042-x86ia64io.h
> +++ b/drivers/input/serio/i8042-x86ia64io.h
> @@ -1107,6 +1107,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>  		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>  					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>  	},
> +	{
> +		/*
> +		 * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
> +		 * the keyboard very laggy for ~5 seconds after boot and
> +		 * sometimes also after resume.
> +		 * However both are required for the keyboard to not fail
> +		 * completely sometimes after boot or resume.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_BOARD_NAME, "N150CU"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
>  	{
>  		.matches = {
>  			DMI_MATCH(DMI_BOARD_NAME, "NH5xAx"),
> @@ -1114,6 +1128,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>  		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>  					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>  	},
> +	{
> +		/*
> +		 * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
> +		 * the keyboard very laggy for ~5 seconds after boot and
> +		 * sometimes also after resume.
> +		 * However both are required for the keyboard to not fail
> +		 * completely sometimes after boot or resume.
> +		 */
> +		.matches = {
> +			DMI_MATCH(DMI_BOARD_NAME, "NHxxRZQ"),
> +		},
> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
> +	},
>  	{
>  		.matches = {
>  			DMI_MATCH(DMI_BOARD_NAME, "NL5xRU"),

Re: [PATCH 1/2] Input: i8042 - add additional TUXEDO devices to i8042 quirk tables

[Dmitry Torokhov]

On Fri, Jul 08, 2022 at 06:10:04PM +0200, Werner Sembach wrote:
> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
> have an external PS/2 port so this can safely be set for all of them.
> 
> I'm not entirely sure if every device listed really needs all four quirks,
> but after testing and production use. No negative effects could be
> observed when setting all four.
> 
> Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
> Cc: stable@vger.kernel.org

Applied, thank you.

-- 
Dmitry

Re: [PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Werner Sembach]

Hi,

On 7/8/22 21:39, Hans de Goede wrote:
> Hi,
>
> On 7/8/22 18:10, Werner Sembach wrote:
>> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
>> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
>> have an external PS/2 port so this can safely be set for all of them.
>>
>> I'm not entirely sure if every device listed really needs all four quirks,
>> but after testing and production use. No negative effects could be
>> observed when setting all four.
>>
>> Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
>> and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
>> boot and sometimes also after resume. However both are required for the
>> keyboard to not fail completely sometimes after boot or resume.
> Hmm, the very laggy bit does not sound good. Have you looked into other
> solutions, e.g. what happens if you use just nomux without any of the
> other 3 options ?

I tried a lot of combinations, but it was some time ago.

iirc: at least nomux and reset are required and both individually cause 
the lagging.

So the issue is not fixed by just using a different set of quirks.

Regards,

Werner

>
> Regards,
>
> Hans
>
>
>> Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
>> Cc: stable@vger.kernel.org
>> ---
>>   drivers/input/serio/i8042-x86ia64io.h | 28 +++++++++++++++++++++++++++
>>   1 file changed, 28 insertions(+)
>>
>> diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
>> index 5204a7dd61d4..9dc0266e5168 100644
>> --- a/drivers/input/serio/i8042-x86ia64io.h
>> +++ b/drivers/input/serio/i8042-x86ia64io.h
>> @@ -1107,6 +1107,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>>   		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>   					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>   	},
>> +	{
>> +		/*
>> +		 * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
>> +		 * the keyboard very laggy for ~5 seconds after boot and
>> +		 * sometimes also after resume.
>> +		 * However both are required for the keyboard to not fail
>> +		 * completely sometimes after boot or resume.
>> +		 */
>> +		.matches = {
>> +			DMI_MATCH(DMI_BOARD_NAME, "N150CU"),
>> +		},
>> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>> +	},
>>   	{
>>   		.matches = {
>>   			DMI_MATCH(DMI_BOARD_NAME, "NH5xAx"),
>> @@ -1114,6 +1128,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>>   		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>   					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>   	},
>> +	{
>> +		/*
>> +		 * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
>> +		 * the keyboard very laggy for ~5 seconds after boot and
>> +		 * sometimes also after resume.
>> +		 * However both are required for the keyboard to not fail
>> +		 * completely sometimes after boot or resume.
>> +		 */
>> +		.matches = {
>> +			DMI_MATCH(DMI_BOARD_NAME, "NHxxRZQ"),
>> +		},
>> +		.driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>> +					SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>> +	},
>>   	{
>>   		.matches = {
>>   			DMI_MATCH(DMI_BOARD_NAME, "NL5xRU"),

Re: [PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Hans de Goede]

Hi,

On 7/11/22 14:45, Werner Sembach wrote:
> Hi,
> 
> On 7/8/22 21:39, Hans de Goede wrote:
>> Hi,
>>
>> On 7/8/22 18:10, Werner Sembach wrote:
>>> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
>>> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
>>> have an external PS/2 port so this can safely be set for all of them.
>>>
>>> I'm not entirely sure if every device listed really needs all four quirks,
>>> but after testing and production use. No negative effects could be
>>> observed when setting all four.
>>>
>>> Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
>>> and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
>>> boot and sometimes also after resume. However both are required for the
>>> keyboard to not fail completely sometimes after boot or resume.
>> Hmm, the very laggy bit does not sound good. Have you looked into other
>> solutions, e.g. what happens if you use just nomux without any of the
>> other 3 options ?
> 
> I tried a lot of combinations, but it was some time ago.
> 
> iirc: at least nomux and reset are required and both individually cause the lagging.
> 
> So the issue is not fixed by just using a different set of quirks.

Hmm, ok. So given that this seems to be the best we can do
the patch looks good to me:

Reviewed-by: Hans de Goede <hdegoede@redhat.com>

Regards,

Hans



>>> Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
>>> Cc: stable@vger.kernel.org
>>> ---
>>>   drivers/input/serio/i8042-x86ia64io.h | 28 +++++++++++++++++++++++++++
>>>   1 file changed, 28 insertions(+)
>>>
>>> diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
>>> index 5204a7dd61d4..9dc0266e5168 100644
>>> --- a/drivers/input/serio/i8042-x86ia64io.h
>>> +++ b/drivers/input/serio/i8042-x86ia64io.h
>>> @@ -1107,6 +1107,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>>>           .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>>                       SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>>       },
>>> +    {
>>> +        /*
>>> +         * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
>>> +         * the keyboard very laggy for ~5 seconds after boot and
>>> +         * sometimes also after resume.
>>> +         * However both are required for the keyboard to not fail
>>> +         * completely sometimes after boot or resume.
>>> +         */
>>> +        .matches = {
>>> +            DMI_MATCH(DMI_BOARD_NAME, "N150CU"),
>>> +        },
>>> +        .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>> +                    SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>> +    },
>>>       {
>>>           .matches = {
>>>               DMI_MATCH(DMI_BOARD_NAME, "NH5xAx"),
>>> @@ -1114,6 +1128,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>>>           .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>>                       SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>>       },
>>> +    {
>>> +        /*
>>> +         * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
>>> +         * the keyboard very laggy for ~5 seconds after boot and
>>> +         * sometimes also after resume.
>>> +         * However both are required for the keyboard to not fail
>>> +         * completely sometimes after boot or resume.
>>> +         */
>>> +        .matches = {
>>> +            DMI_MATCH(DMI_BOARD_NAME, "NHxxRZQ"),
>>> +        },
>>> +        .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>> +                    SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>> +    },
>>>       {
>>>           .matches = {
>>>               DMI_MATCH(DMI_BOARD_NAME, "NL5xRU"),
> 

Re: [PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Werner Sembach]

Am 11.07.22 um 14:55 schrieb Hans de Goede:
> Hi,
>
> On 7/11/22 14:45, Werner Sembach wrote:
>> Hi,
>>
>> On 7/8/22 21:39, Hans de Goede wrote:
>>> Hi,
>>>
>>> On 7/8/22 18:10, Werner Sembach wrote:
>>>> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
>>>> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
>>>> have an external PS/2 port so this can safely be set for all of them.
>>>>
>>>> I'm not entirely sure if every device listed really needs all four quirks,
>>>> but after testing and production use. No negative effects could be
>>>> observed when setting all four.
>>>>
>>>> Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
>>>> and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
>>>> boot and sometimes also after resume. However both are required for the
>>>> keyboard to not fail completely sometimes after boot or resume.
>>> Hmm, the very laggy bit does not sound good. Have you looked into other
>>> solutions, e.g. what happens if you use just nomux without any of the
>>> other 3 options ?
>> I tried a lot of combinations, but it was some time ago.
>>
>> iirc: at least nomux and reset are required and both individually cause the lagging.
>>
>> So the issue is not fixed by just using a different set of quirks.
> Hmm, ok. So given that this seems to be the best we can do
> the patch looks good to me:
>
> Reviewed-by: Hans de Goede <hdegoede@redhat.com>
>
> Regards,
>
> Hans

Afaik this patch never got merged. Sadly I still have no better solution, so I 
wanted to bring the patch up for discussion again as it still makes the 
situation better in my opinion.

Kind Regards,

Werner Sembach

>
>
>>>> Signed-off-by: Werner Sembach <wse@tuxedocomputers.com>
>>>> Cc: stable@vger.kernel.org
>>>> ---
>>>>    drivers/input/serio/i8042-x86ia64io.h | 28 +++++++++++++++++++++++++++
>>>>    1 file changed, 28 insertions(+)
>>>>
>>>> diff --git a/drivers/input/serio/i8042-x86ia64io.h b/drivers/input/serio/i8042-x86ia64io.h
>>>> index 5204a7dd61d4..9dc0266e5168 100644
>>>> --- a/drivers/input/serio/i8042-x86ia64io.h
>>>> +++ b/drivers/input/serio/i8042-x86ia64io.h
>>>> @@ -1107,6 +1107,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>>>>            .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>>>                        SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>>>        },
>>>> +    {
>>>> +        /*
>>>> +         * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
>>>> +         * the keyboard very laggy for ~5 seconds after boot and
>>>> +         * sometimes also after resume.
>>>> +         * However both are required for the keyboard to not fail
>>>> +         * completely sometimes after boot or resume.
>>>> +         */
>>>> +        .matches = {
>>>> +            DMI_MATCH(DMI_BOARD_NAME, "N150CU"),
>>>> +        },
>>>> +        .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>>> +                    SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>>> +    },
>>>>        {
>>>>            .matches = {
>>>>                DMI_MATCH(DMI_BOARD_NAME, "NH5xAx"),
>>>> @@ -1114,6 +1128,20 @@ static const struct dmi_system_id i8042_dmi_quirk_table[] __initconst = {
>>>>            .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>>>                        SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>>>        },
>>>> +    {
>>>> +        /*
>>>> +         * Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS makes
>>>> +         * the keyboard very laggy for ~5 seconds after boot and
>>>> +         * sometimes also after resume.
>>>> +         * However both are required for the keyboard to not fail
>>>> +         * completely sometimes after boot or resume.
>>>> +         */
>>>> +        .matches = {
>>>> +            DMI_MATCH(DMI_BOARD_NAME, "NHxxRZQ"),
>>>> +        },
>>>> +        .driver_data = (void *)(SERIO_QUIRK_NOMUX | SERIO_QUIRK_RESET_ALWAYS |
>>>> +                    SERIO_QUIRK_NOLOOP | SERIO_QUIRK_NOPNP)
>>>> +    },
>>>>        {
>>>>            .matches = {
>>>>                DMI_MATCH(DMI_BOARD_NAME, "NL5xRU"),
>

Re: [PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Dmitry Torokhov]

Hi Werner,

On Tue, Mar 21, 2023 at 05:10:19PM +0100, Werner Sembach wrote:
> Am 11.07.22 um 14:55 schrieb Hans de Goede:
> > Hi,
> > 
> > On 7/11/22 14:45, Werner Sembach wrote:
> > > Hi,
> > > 
> > > On 7/8/22 21:39, Hans de Goede wrote:
> > > > Hi,
> > > > 
> > > > On 7/8/22 18:10, Werner Sembach wrote:
> > > > > A lot of modern Clevo barebones have touchpad and/or keyboard issues after
> > > > > suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
> > > > > have an external PS/2 port so this can safely be set for all of them.
> > > > > 
> > > > > I'm not entirely sure if every device listed really needs all four quirks,
> > > > > but after testing and production use. No negative effects could be
> > > > > observed when setting all four.
> > > > > 
> > > > > Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
> > > > > and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
> > > > > boot and sometimes also after resume. However both are required for the
> > > > > keyboard to not fail completely sometimes after boot or resume.
> > > > Hmm, the very laggy bit does not sound good. Have you looked into other
> > > > solutions, e.g. what happens if you use just nomux without any of the
> > > > other 3 options ?
> > > I tried a lot of combinations, but it was some time ago.
> > > 
> > > iirc: at least nomux and reset are required and both individually cause the lagging.
> > > 
> > > So the issue is not fixed by just using a different set of quirks.
> > Hmm, ok. So given that this seems to be the best we can do
> > the patch looks good to me:
> > 
> > Reviewed-by: Hans de Goede <hdegoede@redhat.com>
> > 
> > Regards,
> > 
> > Hans
> 
> Afaik this patch never got merged. Sadly I still have no better solution, so
> I wanted to bring the patch up for discussion again as it still makes the
> situation better in my opinion.

Could you rebase on top of the latest kernel?

Thanks.

-- 
Dmitry

Re: [PATCH 2/2] Input: i8042 - add TUXEDO devices to i8042 quirk tables for partial fix

[Werner Sembach]

Am 21.03.23 um 19:53 schrieb Dmitry Torokhov:
> Hi Werner,
>
> On Tue, Mar 21, 2023 at 05:10:19PM +0100, Werner Sembach wrote:
>> Am 11.07.22 um 14:55 schrieb Hans de Goede:
>>> Hi,
>>>
>>> On 7/11/22 14:45, Werner Sembach wrote:
>>>> Hi,
>>>>
>>>> On 7/8/22 21:39, Hans de Goede wrote:
>>>>> Hi,
>>>>>
>>>>> On 7/8/22 18:10, Werner Sembach wrote:
>>>>>> A lot of modern Clevo barebones have touchpad and/or keyboard issues after
>>>>>> suspend fixable with nomux + reset + noloop + nopnp. Luckily, none of them
>>>>>> have an external PS/2 port so this can safely be set for all of them.
>>>>>>
>>>>>> I'm not entirely sure if every device listed really needs all four quirks,
>>>>>> but after testing and production use. No negative effects could be
>>>>>> observed when setting all four.
>>>>>>
>>>>>> Setting SERIO_QUIRK_NOMUX or SERIO_QUIRK_RESET_ALWAYS on the Clevo N150CU
>>>>>> and the Clevo NHxxRZQ makes the keyboard very laggy for ~5 seconds after
>>>>>> boot and sometimes also after resume. However both are required for the
>>>>>> keyboard to not fail completely sometimes after boot or resume.
>>>>> Hmm, the very laggy bit does not sound good. Have you looked into other
>>>>> solutions, e.g. what happens if you use just nomux without any of the
>>>>> other 3 options ?
>>>> I tried a lot of combinations, but it was some time ago.
>>>>
>>>> iirc: at least nomux and reset are required and both individually cause the lagging.
>>>>
>>>> So the issue is not fixed by just using a different set of quirks.
>>> Hmm, ok. So given that this seems to be the best we can do
>>> the patch looks good to me:
>>>
>>> Reviewed-by: Hans de Goede <hdegoede@redhat.com>
>>>
>>> Regards,
>>>
>>> Hans
>> Afaik this patch never got merged. Sadly I still have no better solution, so
>> I wanted to bring the patch up for discussion again as it still makes the
>> situation better in my opinion.
> Could you rebase on top of the latest kernel?
>
> Thanks.
>
ofc, sent as v3