mbox series

[v2,0/5] thunderbolt / ACPI: Add support for USB4 _OSC

Message ID 20210129083241.72497-1-mika.westerberg@linux.intel.com
Headers show
Series thunderbolt / ACPI: Add support for USB4 _OSC | expand

Message

Mika Westerberg Jan. 29, 2021, 8:32 a.m. UTC
Hi all,

The just released ACPI 6.4 spec [1] adds a new _OSC method that is used to
negotiate OS support for native USB4 features such as PCIe tunneling. This
patch series adds Linux support for the new _OSC and modifies the
Thunderbolt/USB4 driver accordingly to enable/disable tunneling of
different protocols.

There is an additional setting in the firmware connection manager that
allows the BIOS to disable PCIe tunneling, so we add support for this and
also make the software connection manager to switch to this "nopcie"
security level when the _OSC does not allow PCIe tunneling.

This applies on top of thunderbolt.git/next.

[1] https://uefi.org/sites/default/files/resources/ACPI_Spec_6_4_Jan22.pdf

The previous version of the patch series can be found here:

  https://lore.kernel.org/linux-usb/20210126155723.9388-1-mika.westerberg@linux.intel.com/

Changes from v1:

  * Dropped patch 1/6. I already applied it to thunderbolt.git/fixes
  * Added ACK from Yehezkel to TBT patches
  * Updated changelog of patch 1/5 and fixed typos too
  * Added Rafael's tag to patch 4/5.

Mario Limonciello (1):
  ACPI: Execute platform _OSC also with query bit clear

Mika Westerberg (4):
  thunderbolt: Add support for PCIe tunneling disabled (SL5)
  thunderbolt: Allow disabling XDomain protocol
  ACPI: Add support for native USB4 control _OSC
  thunderbolt: Add support for native USB4 _OSC

 .../ABI/testing/sysfs-bus-thunderbolt         |   2 +
 Documentation/admin-guide/thunderbolt.rst     |   7 ++
 drivers/acpi/bus.c                            | 119 ++++++++++++++++--
 drivers/thunderbolt/acpi.c                    |  65 ++++++++++
 drivers/thunderbolt/domain.c                  |  16 ++-
 drivers/thunderbolt/icm.c                     |   6 +-
 drivers/thunderbolt/nhi.c                     |  27 +++-
 drivers/thunderbolt/switch.c                  |   6 +-
 drivers/thunderbolt/tb.c                      |  22 +++-
 drivers/thunderbolt/tb.h                      |  13 ++
 drivers/thunderbolt/tunnel.c                  |  10 +-
 drivers/thunderbolt/usb4.c                    |  11 +-
 drivers/thunderbolt/xdomain.c                 |   9 ++
 include/linux/acpi.h                          |  10 ++
 include/linux/thunderbolt.h                   |   3 +
 15 files changed, 298 insertions(+), 28 deletions(-)

Comments

Mika Westerberg Feb. 3, 2021, 8:14 a.m. UTC | #1
Hi Rafael,

I wonder if you are OK with this patch?

Thanks!

On Fri, Jan 29, 2021 at 11:32:39AM +0300, Mika Westerberg wrote:
> From: Mario Limonciello <mario.limonciello@dell.com>

> 

> The platform _OSC can change the hardware state when query bit is not

> set. According to ACPI spec it is recommended that the OS runs _OSC with

> query bit set until the platform does not mask any of the capabilities.

> Then it should run it with query bit clear in order to actually commit

> the changes. Linux has not been doing this for the reasons that there

> has not been anything to commit, until now.

> 

> The ACPI 6.4 introduced _OSC for USB4 to allow the OS to negotiate

> native control over USB4 tunneling. The platform might implement this so

> that it only activates the software connection manager path when the OS

> calls the _OSC with the query bit clear. Otherwise it may default to the

> firmware connection manager, for instance.

> 

> For this reason modify the _OSC support so that we first execute it with

> query bit set, then use the returned value as base of the features we

> want to control and run the _OSC again with query bit clear. This also

> follows what Windows is doing.

> 

> Also rename the function to better match what it does.

> 

> Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

> Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>

> Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>

> ---

>  drivers/acpi/bus.c | 43 +++++++++++++++++++++++++++++++------------

>  1 file changed, 31 insertions(+), 12 deletions(-)

> 

> diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c

> index 1682f8b454a2..a52cb28c40d8 100644

> --- a/drivers/acpi/bus.c

> +++ b/drivers/acpi/bus.c

> @@ -282,9 +282,9 @@ bool osc_pc_lpi_support_confirmed;

>  EXPORT_SYMBOL_GPL(osc_pc_lpi_support_confirmed);

>  

>  static u8 sb_uuid_str[] = "0811B06E-4A27-44F9-8D60-3CBBC22E7B48";

> -static void acpi_bus_osc_support(void)

> +static void acpi_bus_osc_negotiate_platform_control(void)

>  {

> -	u32 capbuf[2];

> +	u32 capbuf[2], *capbuf_ret;

>  	struct acpi_osc_context context = {

>  		.uuid_str = sb_uuid_str,

>  		.rev = 1,

> @@ -321,17 +321,36 @@ static void acpi_bus_osc_support(void)

>  		capbuf[OSC_SUPPORT_DWORD] |= OSC_SB_APEI_SUPPORT;

>  	if (ACPI_FAILURE(acpi_get_handle(NULL, "\\_SB", &handle)))

>  		return;

> -	if (ACPI_SUCCESS(acpi_run_osc(handle, &context))) {

> -		u32 *capbuf_ret = context.ret.pointer;

> -		if (context.ret.length > OSC_SUPPORT_DWORD) {

> -			osc_sb_apei_support_acked =

> -				capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_APEI_SUPPORT;

> -			osc_pc_lpi_support_confirmed =

> -				capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_PCLPI_SUPPORT;

> -		}

> +

> +	if (ACPI_FAILURE(acpi_run_osc(handle, &context)))

> +		return;

> +

> +	capbuf_ret = context.ret.pointer;

> +	if (context.ret.length <= OSC_SUPPORT_DWORD) {

>  		kfree(context.ret.pointer);

> +		return;

>  	}

> -	/* do we need to check other returned cap? Sounds no */

> +

> +	/*

> +	 * Now run _OSC again with query flag clear and with the caps

> +	 * supported by both the OS and the platform.

> +	 */

> +	capbuf[OSC_QUERY_DWORD] = 0;

> +	capbuf[OSC_SUPPORT_DWORD] = capbuf_ret[OSC_SUPPORT_DWORD];

> +	kfree(context.ret.pointer);

> +

> +	if (ACPI_FAILURE(acpi_run_osc(handle, &context)))

> +		return;

> +

> +	capbuf_ret = context.ret.pointer;

> +	if (context.ret.length > OSC_SUPPORT_DWORD) {

> +		osc_sb_apei_support_acked =

> +			capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_APEI_SUPPORT;

> +		osc_pc_lpi_support_confirmed =

> +			capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_PCLPI_SUPPORT;

> +	}

> +

> +	kfree(context.ret.pointer);

>  }

>  

>  /* --------------------------------------------------------------------------

> @@ -1168,7 +1187,7 @@ static int __init acpi_bus_init(void)

>  	 * _OSC method may exist in module level code,

>  	 * so it must be run after ACPI_FULL_INITIALIZATION

>  	 */

> -	acpi_bus_osc_support();

> +	acpi_bus_osc_negotiate_platform_control();

>  

>  	/*

>  	 * _PDC control method may load dynamic SSDT tables,

> -- 

> 2.29.2
Mika Westerberg Feb. 4, 2021, 7:36 a.m. UTC | #2
On Wed, Feb 03, 2021 at 02:56:25PM +0100, Rafael J. Wysocki wrote:
> On Wed, Feb 3, 2021 at 9:16 AM Mika Westerberg
> <mika.westerberg@linux.intel.com> wrote:
> >
> > Hi Rafael,
> >
> > I wonder if you are OK with this patch?
> 
> It looks good to me now, please feel free to add my ACK to it.

Thanks!
Mika Westerberg Feb. 4, 2021, 7:51 a.m. UTC | #3
On Fri, Jan 29, 2021 at 11:32:36AM +0300, Mika Westerberg wrote:
> Hi all,
> 
> The just released ACPI 6.4 spec [1] adds a new _OSC method that is used to
> negotiate OS support for native USB4 features such as PCIe tunneling. This
> patch series adds Linux support for the new _OSC and modifies the
> Thunderbolt/USB4 driver accordingly to enable/disable tunneling of
> different protocols.
> 
> There is an additional setting in the firmware connection manager that
> allows the BIOS to disable PCIe tunneling, so we add support for this and
> also make the software connection manager to switch to this "nopcie"
> security level when the _OSC does not allow PCIe tunneling.
> 
> This applies on top of thunderbolt.git/next.
> 
> [1] https://uefi.org/sites/default/files/resources/ACPI_Spec_6_4_Jan22.pdf
> 
> The previous version of the patch series can be found here:
> 
>   https://lore.kernel.org/linux-usb/20210126155723.9388-1-mika.westerberg@linux.intel.com/
> 
> Changes from v1:
> 
>   * Dropped patch 1/6. I already applied it to thunderbolt.git/fixes
>   * Added ACK from Yehezkel to TBT patches
>   * Updated changelog of patch 1/5 and fixed typos too
>   * Added Rafael's tag to patch 4/5.
> 
> Mario Limonciello (1):
>   ACPI: Execute platform _OSC also with query bit clear
> 
> Mika Westerberg (4):
>   thunderbolt: Add support for PCIe tunneling disabled (SL5)
>   thunderbolt: Allow disabling XDomain protocol
>   ACPI: Add support for native USB4 control _OSC
>   thunderbolt: Add support for native USB4 _OSC

All applied to thunderbolt.git/next.
Lee, Chun-Yi June 7, 2021, 12:31 p.m. UTC | #4
Hi Mika,

There have some machines be found on openSUSE Tumbleweed that this patch
causes that SSDT tables can not be dynamic loaded. The symptom is that
dmesg shows '_CPC not found' because SSDT table did not dynamic load.

[    1.149107] ACPI BIOS Error (bug): Could not resolve symbol [\_PR.PR00._CPC], AE_NOT_FOUND (20210105/psargs-330)

Looks that the firmware didn't response OSC_SB_CPCV2_SUPPORT after
kernel changed to new behavior. The openSUSE bug is here:

Bug 1185513 - ACPI BIOS Error after upgrade to 5.12.0-1-default 
https://bugzilla.suse.com/show_bug.cgi?id=1185513

Could you please help to give any suggestion?

Thanks a lot!
Joey Lee

On Wed, Feb 03, 2021 at 10:14:15AM +0200, Mika Westerberg wrote:
> Hi Rafael,

> 

> I wonder if you are OK with this patch?

> 

> Thanks!

> 

> On Fri, Jan 29, 2021 at 11:32:39AM +0300, Mika Westerberg wrote:

> > From: Mario Limonciello <mario.limonciello@dell.com>

> > 

> > The platform _OSC can change the hardware state when query bit is not

> > set. According to ACPI spec it is recommended that the OS runs _OSC with

> > query bit set until the platform does not mask any of the capabilities.

> > Then it should run it with query bit clear in order to actually commit

> > the changes. Linux has not been doing this for the reasons that there

> > has not been anything to commit, until now.

> > 

> > The ACPI 6.4 introduced _OSC for USB4 to allow the OS to negotiate

> > native control over USB4 tunneling. The platform might implement this so

> > that it only activates the software connection manager path when the OS

> > calls the _OSC with the query bit clear. Otherwise it may default to the

> > firmware connection manager, for instance.

> > 

> > For this reason modify the _OSC support so that we first execute it with

> > query bit set, then use the returned value as base of the features we

> > want to control and run the _OSC again with query bit clear. This also

> > follows what Windows is doing.

> > 

> > Also rename the function to better match what it does.

> > 

> > Cc: Rafael J. Wysocki <rafael.j.wysocki@intel.com>

> > Signed-off-by: Mario Limonciello <mario.limonciello@dell.com>

> > Signed-off-by: Mika Westerberg <mika.westerberg@linux.intel.com>

> > ---

> >  drivers/acpi/bus.c | 43 +++++++++++++++++++++++++++++++------------

> >  1 file changed, 31 insertions(+), 12 deletions(-)

> > 

> > diff --git a/drivers/acpi/bus.c b/drivers/acpi/bus.c

> > index 1682f8b454a2..a52cb28c40d8 100644

> > --- a/drivers/acpi/bus.c

> > +++ b/drivers/acpi/bus.c

> > @@ -282,9 +282,9 @@ bool osc_pc_lpi_support_confirmed;

> >  EXPORT_SYMBOL_GPL(osc_pc_lpi_support_confirmed);

> >  

> >  static u8 sb_uuid_str[] = "0811B06E-4A27-44F9-8D60-3CBBC22E7B48";

> > -static void acpi_bus_osc_support(void)

> > +static void acpi_bus_osc_negotiate_platform_control(void)

> >  {

> > -	u32 capbuf[2];

> > +	u32 capbuf[2], *capbuf_ret;

> >  	struct acpi_osc_context context = {

> >  		.uuid_str = sb_uuid_str,

> >  		.rev = 1,

> > @@ -321,17 +321,36 @@ static void acpi_bus_osc_support(void)

> >  		capbuf[OSC_SUPPORT_DWORD] |= OSC_SB_APEI_SUPPORT;

> >  	if (ACPI_FAILURE(acpi_get_handle(NULL, "\\_SB", &handle)))

> >  		return;

> > -	if (ACPI_SUCCESS(acpi_run_osc(handle, &context))) {

> > -		u32 *capbuf_ret = context.ret.pointer;

> > -		if (context.ret.length > OSC_SUPPORT_DWORD) {

> > -			osc_sb_apei_support_acked =

> > -				capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_APEI_SUPPORT;

> > -			osc_pc_lpi_support_confirmed =

> > -				capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_PCLPI_SUPPORT;

> > -		}

> > +

> > +	if (ACPI_FAILURE(acpi_run_osc(handle, &context)))

> > +		return;

> > +

> > +	capbuf_ret = context.ret.pointer;

> > +	if (context.ret.length <= OSC_SUPPORT_DWORD) {

> >  		kfree(context.ret.pointer);

> > +		return;

> >  	}

> > -	/* do we need to check other returned cap? Sounds no */

> > +

> > +	/*

> > +	 * Now run _OSC again with query flag clear and with the caps

> > +	 * supported by both the OS and the platform.

> > +	 */

> > +	capbuf[OSC_QUERY_DWORD] = 0;

> > +	capbuf[OSC_SUPPORT_DWORD] = capbuf_ret[OSC_SUPPORT_DWORD];

> > +	kfree(context.ret.pointer);

> > +

> > +	if (ACPI_FAILURE(acpi_run_osc(handle, &context)))

> > +		return;

> > +

> > +	capbuf_ret = context.ret.pointer;

> > +	if (context.ret.length > OSC_SUPPORT_DWORD) {

> > +		osc_sb_apei_support_acked =

> > +			capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_APEI_SUPPORT;

> > +		osc_pc_lpi_support_confirmed =

> > +			capbuf_ret[OSC_SUPPORT_DWORD] & OSC_SB_PCLPI_SUPPORT;

> > +	}

> > +

> > +	kfree(context.ret.pointer);

> >  }

> >  

> >  /* --------------------------------------------------------------------------

> > @@ -1168,7 +1187,7 @@ static int __init acpi_bus_init(void)

> >  	 * _OSC method may exist in module level code,

> >  	 * so it must be run after ACPI_FULL_INITIALIZATION

> >  	 */

> > -	acpi_bus_osc_support();

> > +	acpi_bus_osc_negotiate_platform_control();

> >  

> >  	/*

> >  	 * _PDC control method may load dynamic SSDT tables,

> > -- 

> > 2.29.2
Mika Westerberg June 7, 2021, 4:10 p.m. UTC | #5
Hi,

On Mon, Jun 07, 2021 at 08:31:10PM +0800, joeyli wrote:
> Hi Mika,
> 
> There have some machines be found on openSUSE Tumbleweed that this patch
> causes that SSDT tables can not be dynamic loaded. The symptom is that
> dmesg shows '_CPC not found' because SSDT table did not dynamic load.
> 
> [    1.149107] ACPI BIOS Error (bug): Could not resolve symbol [\_PR.PR00._CPC], AE_NOT_FOUND (20210105/psargs-330)
> 
> Looks that the firmware didn't response OSC_SB_CPCV2_SUPPORT after
> kernel changed to new behavior. The openSUSE bug is here:
> 
> Bug 1185513 - ACPI BIOS Error after upgrade to 5.12.0-1-default 
> https://bugzilla.suse.com/show_bug.cgi?id=1185513
> 
> Could you please help to give any suggestion?

There is another one that Red Hat reported here:

https://bugzilla.kernel.org/show_bug.cgi?id=213023

The Bugzilla entry also has a patch attached [1] from Hans, can you try it
out and see if that fixes the issue?

[1] https://bugzilla.kernel.org/attachment.cgi?id=297195

Thanks!
Lee, Chun-Yi June 8, 2021, 3:02 a.m. UTC | #6
Hi Mika,

On Mon, Jun 07, 2021 at 07:10:59PM +0300, Mika Westerberg wrote:
> Hi,
> 
> On Mon, Jun 07, 2021 at 08:31:10PM +0800, joeyli wrote:
> > Hi Mika,
> > 
> > There have some machines be found on openSUSE Tumbleweed that this patch
> > causes that SSDT tables can not be dynamic loaded. The symptom is that
> > dmesg shows '_CPC not found' because SSDT table did not dynamic load.
> > 
> > [    1.149107] ACPI BIOS Error (bug): Could not resolve symbol [\_PR.PR00._CPC], AE_NOT_FOUND (20210105/psargs-330)
> > 
> > Looks that the firmware didn't response OSC_SB_CPCV2_SUPPORT after
> > kernel changed to new behavior. The openSUSE bug is here:
> > 
> > Bug 1185513 - ACPI BIOS Error after upgrade to 5.12.0-1-default 
> > https://bugzilla.suse.com/show_bug.cgi?id=1185513
> > 
> > Could you please help to give any suggestion?
> 
> There is another one that Red Hat reported here:
> 
> https://bugzilla.kernel.org/show_bug.cgi?id=213023
> 
> The Bugzilla entry also has a patch attached [1] from Hans, can you try it
> out and see if that fixes the issue?
> 
> [1] https://bugzilla.kernel.org/attachment.cgi?id=297195
>

Thanks for your information! Hans's patch looks reasonable.

I will backport to openSUSE TW and let bug reporter helps to test it. We will
add comment on kernel.org bugzilla.

Thanks a lot!
Joey Lee