hw/pxb-pcie: fix PCI Express hotplug support
PATCH v3 1/1 hw/pxb-pcie: fix PCI Express hotplug support
Add the missing osc method for pxb-pcie devices as APCI spec recommends, see 6.2.10.3 OSC Implementation Example for PCI Host Bridge Devices, ACPI 5.0:
It is recommended that a machine with multiple host bridge devices should report the same capabilities for all host bridges, and also negotiate control of the features described in the Control Field in the same way for all host bridges.
---
Note to maintainer:
Please update ACPI test files.
V2 -> V3:
- Modified comment (Igor)
V1 -> V2:
Addressed Michael S. Tsirkin's comments:
- Added documentation to q35 osc function
- Made _osc serialized. I did not add compat property since it seems guest OSs do not care anyway about the OSC being serialized.
- Kept the SUPP field even if is not used and also left both SUPP and CTRL out of the _osc because all systems I checked and the ACPI spec keep them that way, maybe is some kind of documentation contract.
Thanks,
Marcel
hw/i386/acpi-build.c | 41 +++++++++++++++++++++++++++++++++++------ 1 file changed, 35 insertions(+), 6 deletions(-)
hw/i386/acpi-build.c +35 -6
@@ -1796,7 +1796,7 @@ static void build_piix4_pci_hotplug(Aml *table)aml_append (table, scope); } -static Aml *build_q35_osc_method (void ) +static void build_q35_osc_method (Aml *dev) {Aml *if_ctx;Aml *if_ctx2;
@@ -1805,7 +1805,35 @@ static Aml *build_q35_osc_method(void)Aml *a_cwd1 =aml_name ("CDW1" );Aml *a_ctrl =aml_name ("CTRL" ); - method =aml_method ("_OSC" ,4 ,AML_NOTSERIALIZED ); +/* +* Bits defined in the Support Field provide information +* regarding OS supported features. +* +* This field is not actually used and can be removed, +* however it appears even if unused on most DSDTs. +* +* See: +* Table 6-148 Interpretation of _OSC Support Field, +* Passed in via the 2nd dword in Arg3, APCI 5.0 +*/ +aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); + +/* +* Bits defined in the Control Field are used to submit +* request by the OS for control/handling of the associated feature +* +* See: Table 6-149 Interpretation of _OSC Control Field, +* Passed in via Arg3, ACPI 5.0 +* Table 6-150 Interpretation of _OSC Control Field, +* Returned Value, APCI 5.0 +*/ +aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 ))); + + +/* +* 6.2.10 _OSC (Operating System Capabilities), APCI 5.0 +*/ + method =aml_method ("_OSC" ,4 ,AML_SERIALIZED );aml_append (method,aml_create_dword_field (aml_arg (3 ),aml_int (0 ),"CDW1" )); if_ctx = aml_if(aml_equal(
@@ -1842,7 +1870,7 @@ static Aml *build_q35_osc_method(void)aml_append (method, else_ctx);aml_append (method,aml_return (aml_arg (3 ))); -return method; +aml_append (dev, method); } static void
@@ -1898,9 +1926,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_CID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_ADR" ,aml_int (0 )));aml_append (dev,aml_name_decl ("_UID" ,aml_int (1 ))); -aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); -aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 ))); -aml_append (dev,build_q35_osc_method ()); +build_q35_osc_method (dev);aml_append (sb_scope, dev);aml_append (dsdt, sb_scope);
@@ -1964,6 +1990,9 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_UID" ,aml_int (bus_num)));aml_append (dev,aml_name_decl ("_HID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_BBN" ,aml_int (bus_num))); +if (pci_bus_is_express (bus)) { +build_q35_osc_method (dev); + }if (numa_node !=NUMA_NODE_UNASSIGNED ) {aml_append (dev,aml_name_decl ("_PXM" ,aml_int (numa_node)));
Signature · 1 line
2.5.5
Quoted 41 lines · Marcel Apfelbaum
Add the missing osc method for pxb-pcie devices as APCI spec recommends, see 6.2.10.3 OSC Implementation Example for PCI Host Bridge Devices, ACPI 5.0:
It is recommended that a machine with multiple host bridge devices should report the same capabilities for all host bridges, and also negotiate control of the features described in the Control Field in the same way for all host bridges.
Reviewed-by Igor Mammedov <[email protected]>Signed-off-by Marcel Apfelbaum <[email protected]>---
Note to maintainer:
Please update ACPI test files.V2 -> V3:
- Modified comment (Igor)V1 -> V2:
Addressed Michael S. Tsirkin's comments:
- Added documentation to q35 osc function
- Made_oscserialized. I did not add compat property since it seems guest OSs do not care anyway about the OSC being serialized.
- Kept the SUPP field even if is not used and also left both SUPP and CTRL out of the_oscbecause all systems I checked and the ACPI spec keep them that way, maybe is some kind of documentation contract.Thanks,
Marcelhw/i386/acpi-build.c | 41 +++++++++++++++++++++++++++++++++++------ 1 file changed, 35 insertions(+), 6 deletions(-)hw/i386/acpi-build.c +0
-0
hw/i386/acpi-build.c static void build_piix4_pci_hotplug(Aml *table) · line 1799
Quoted 5 lines
hw/i386/acpi-build.c +1
-1@@ -1796,7 +1796,7 @@ static void build_piix4_pci_hotplug(Aml *table)aml_append (table, scope); } -static Aml *build_q35_osc_method (void ) +static void build_q35_osc_method (Aml *dev)
This does more than just build _OSC so please
give it a better name.
Quoted 4 lines
hw/i386/acpi-build.c +0
-0{Aml *if_ctx;Aml *if_ctx2;
hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1813
Quoted 10 lines
hw/i386/acpi-build.c +6
-1@@ -1805,7 +1805,35 @@ static Aml *build_q35_osc_method(void)Aml *a_cwd1 =aml_name ("CDW1" );Aml *a_ctrl =aml_name ("CTRL" ); - method =aml_method ("_OSC" ,4 ,AML_NOTSERIALIZED ); +/* +* Bits defined in the Support Field provide information +* regarding OS supported features. +* +* This field is not actually used and can be removed, +* however it appears even if unused on most DSDTs.
What does this mean? Our _OSC method uses these, that's why we have them. How can you just remove them?
I think you mean we could remove the code writing to it. I don't think keeping it around is justified. Same applies to CTRL btw. We could use a local variable instead.
hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1817
Quoted 4 lines
hw/i386/acpi-build.c +4
-0+* +* See: +* Table 6-148 Interpretation of _OSC Support Field, +* Passed in via the 2nd dword in Arg3, APCI 5.0
Why 5.0? This is not how we do it. Pls find the earliest spec version that has this. 3.0?
Quoted 8 lines
hw/i386/acpi-build.c +8
-0+*/ +aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); + +/* +* Bits defined in the Control Field are used to submit +* request by the OS for control/handling of the associated feature +* +* See: Table 6-149 Interpretation of _OSC Control Field,
hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1835
Quoted 10 lines
hw/i386/acpi-build.c +10
-0+* Passed in via Arg3, ACPI 5.0 +* Table 6-150 Interpretation of _OSC Control Field, +* Returned Value, APCI 5.0 +*/ +aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 ))); + + +/* +* 6.2.10 _OSC (Operating System Capabilities), APCI 5.0 +*/
Same here.
hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1873
Quoted 10 lines
hw/i386/acpi-build.c +2
-1+ method =aml_method ("_OSC" ,4 ,AML_SERIALIZED );aml_append (method,aml_create_dword_field (aml_arg (3 ),aml_int (0 ),"CDW1" )); if_ctx = aml_if(aml_equal(@@ -1842,7 +1870,7 @@ static Aml *build_q35_osc_method(void)aml_append (method, else_ctx);aml_append (method,aml_return (aml_arg (3 ))); -return method; +aml_append (dev, method);
If you drop SUPP and CTRL then you can keep it simple, right?
Quoted 26 lines
hw/i386/acpi-build.c +4
-3} static void@@ -1898,9 +1926,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_CID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_ADR" ,aml_int (0 )));aml_append (dev,aml_name_decl ("_UID" ,aml_int (1 ))); -aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); -aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 ))); -aml_append (dev,build_q35_osc_method ()); +build_q35_osc_method (dev);aml_append (sb_scope, dev);aml_append (dsdt, sb_scope);@@ -1964,6 +1990,9 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_UID" ,aml_int (bus_num)));aml_append (dev,aml_name_decl ("_HID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_BBN" ,aml_int (bus_num))); +if (pci_bus_is_express (bus)) { +build_q35_osc_method (dev); + }if (numa_node !=NUMA_NODE_UNASSIGNED ) {aml_append (dev,aml_name_decl ("_PXM" ,aml_int (numa_node)));Signature · 1 line
2.5.5
so how about
--->
acpi: simplify _OSC
Our _OSC method has a bunch of unused code loading data into external CTRL and SUPP fields which are then never used. Drop this in favor of a single local variable.
---
hw/i386/acpi-build.c +1 -4
@@ -1804,7 +1804,7 @@ static Aml *build_q35_osc_method(void)Aml *else_ctx;Aml *method;Aml *a_cwd1 =aml_name ("CDW1" ); -Aml *a_ctrl =aml_name ("CTRL" ); +Aml *a_ctrl =aml_local (0 ); method =aml_method ("_OSC" ,4 ,AML_NOTSERIALIZED );aml_append (method,aml_create_dword_field (aml_arg (3 ),aml_int (0 ),"CDW1" ));
@@ -1814,7 +1814,6 @@ static Aml *build_q35_osc_method(void)aml_append (if_ctx,aml_create_dword_field (aml_arg (3 ),aml_int (4 ),"CDW2" ));aml_append (if_ctx,aml_create_dword_field (aml_arg (3 ),aml_int (8 ),"CDW3" )); -aml_append (if_ctx,aml_store (aml_name ("CDW2" ),aml_name ("SUPP" )));aml_append (if_ctx,aml_store (aml_name ("CDW3" ), a_ctrl));/*
@@ -1899,8 +1898,6 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_CID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_ADR" ,aml_int (0 )));aml_append (dev,aml_name_decl ("_UID" ,aml_int (1 ))); -aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); -aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 )));aml_append (dev,build_q35_osc_method ());aml_append (sb_scope, dev);aml_append (dsdt, sb_scope);
Quoted 182 lines · Michael S. Tsirkin
Quoted 41 lines · Marcel Apfelbaum
Add the missing osc method for pxb-pcie devices as APCI spec recommends, see 6.2.10.3 OSC Implementation Example for PCI Host Bridge Devices, ACPI 5.0:
It is recommended that a machine with multiple host bridge devices should report the same capabilities for all host bridges, and also negotiate control of the features described in the Control Field in the same way for all host bridges.
Reviewed-by Igor Mammedov <[email protected]>Signed-off-by Marcel Apfelbaum <[email protected]>---
Note to maintainer:
Please update ACPI test files.V2 -> V3:
- Modified comment (Igor)V1 -> V2:
Addressed Michael S. Tsirkin's comments:
- Added documentation to q35 osc function
- Made_oscserialized. I did not add compat property since it seems guest OSs do not care anyway about the OSC being serialized.
- Kept the SUPP field even if is not used and also left both SUPP and CTRL out of the_oscbecause all systems I checked and the ACPI spec keep them that way, maybe is some kind of documentation contract.Thanks,
Marcelhw/i386/acpi-build.c | 41 +++++++++++++++++++++++++++++++++++------ 1 file changed, 35 insertions(+), 6 deletions(-)hw/i386/acpi-build.c +0
-0hw/i386/acpi-build.c static void build_piix4_pci_hotplug(Aml *table) · line 1799
Quoted 5 lines
hw/i386/acpi-build.c +1
-1@@ -1796,7 +1796,7 @@ static void build_piix4_pci_hotplug(Aml *table)aml_append (table, scope); } -static Aml *build_q35_osc_method (void ) +static void build_q35_osc_method (Aml *dev)This does more than just build
_OSCso please
give it a better name.Quoted 4 lines
hw/i386/acpi-build.c +0
-0{Aml *if_ctx;Aml *if_ctx2;hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1813
Quoted 10 lines
hw/i386/acpi-build.c +6
-1@@ -1805,7 +1805,35 @@ static Aml *build_q35_osc_method(void)Aml *a_cwd1 =aml_name ("CDW1" );Aml *a_ctrl =aml_name ("CTRL" ); - method =aml_method ("_OSC" ,4 ,AML_NOTSERIALIZED ); +/* +* Bits defined in the Support Field provide information +* regarding OS supported features. +* +* This field is not actually used and can be removed, +* however it appears even if unused on most DSDTs. What does this mean? Our
_OSCmethod uses these, that's why we have them. How can you just remove them?I think you mean we could remove the code writing to it. I don't think keeping it around is justified. Same applies to CTRL btw. We could use a local variable instead.
hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1817
Quoted 4 lines
hw/i386/acpi-build.c +4
-0+* +* See: +* Table 6-148 Interpretation of _OSC Support Field, +* Passed in via the 2nd dword in Arg3, APCI 5.0 Why 5.0? This is not how we do it. Pls find the earliest spec version that has this. 3.0?
Quoted 8 lines
hw/i386/acpi-build.c +8
-0+*/ +aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); + +/* +* Bits defined in the Control Field are used to submit +* request by the OS for control/handling of the associated feature +* +* See: Table 6-149 Interpretation of _OSC Control Field, hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1835
Quoted 10 lines
hw/i386/acpi-build.c +10
-0+* Passed in via Arg3, ACPI 5.0 +* Table 6-150 Interpretation of _OSC Control Field, +* Returned Value, APCI 5.0 +*/ +aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 ))); + + +/* +* 6.2.10 _OSC (Operating System Capabilities), APCI 5.0 +*/ Same here.
hw/i386/acpi-build.c static Aml *build_q35_osc_method(void) · line 1873
Quoted 10 lines
hw/i386/acpi-build.c +2
-1+ method =aml_method ("_OSC" ,4 ,AML_SERIALIZED );aml_append (method,aml_create_dword_field (aml_arg (3 ),aml_int (0 ),"CDW1" )); if_ctx = aml_if(aml_equal(@@ -1842,7 +1870,7 @@ static Aml *build_q35_osc_method(void)aml_append (method, else_ctx);aml_append (method,aml_return (aml_arg (3 ))); -return method; +aml_append (dev, method);If you drop SUPP and CTRL then you can keep it simple, right?
Quoted 25 lines
hw/i386/acpi-build.c +4
-4} static void@@ -1898,9 +1926,7 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_CID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_ADR" ,aml_int (0 )));aml_append (dev,aml_name_decl ("_UID" ,aml_int (1 ))); -aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); -aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 ))); -aml_append (dev,build_q35_osc_method ()); +build_q35_osc_method (dev);aml_append (sb_scope, dev);aml_append (dsdt, sb_scope);@@ -1964,6 +1990,9 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_UID" ,aml_int (bus_num)));aml_append (dev,aml_name_decl ("_HID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_BBN" ,aml_int (bus_num))); +if (pci_bus_is_express (bus)) { +build_q35_osc_method (dev); + }if (numa_node !=NUMA_NODE_UNASSIGNED ) {aml_append (dev,aml_name_decl ("_PXM" ,aml_int (numa_node))); --Quoted 1 line
2.5.5
so how about
--->
acpi: simplify
_OSCOur
_OSCmethod has a bunch of unused code loading data into external CTRL and SUPP fields which are then never used. Drop this in favor of a single local variable.Signed-off-by Michael S. Tsirkin <[email protected]>---
hw/i386/acpi-build.c +1
-2@@ -1804,7 +1804,7 @@ static Aml *build_q35_osc_method(void)Aml *else_ctx;Aml *method;Aml *a_cwd1 =aml_name ("CDW1" ); -Aml *a_ctrl =aml_name ("CTRL" ); +Aml *a_ctrl =aml_local (0 ); method =aml_method ("_OSC" ,4 ,AML_NOTSERIALIZED );aml_append (method,aml_create_dword_field (aml_arg (3 ),aml_int (0 ),"CDW1" ));@@ -1814,7 +1814,6 @@ static Aml *build_q35_osc_method(void)aml_append (if_ctx,aml_create_dword_field (aml_arg (3 ),aml_int (4 ),"CDW2" ));aml_append (if_ctx,aml_create_dword_field (aml_arg (3 ),aml_int (8 ),"CDW3" )); -aml_append (if_ctx,aml_store (aml_name ("CDW2" ),aml_name ("SUPP" )));aml_append (if_ctx,aml_store (aml_name ("CDW3" ), a_ctrl));/*
hw/i386/acpi-build.c build_dsdt(GArray *table_data, BIOSLinker *linker, · line 1903
Quoted 8 lines
hw/i386/acpi-build.c +0
-2@@ -1899,8 +1898,6 @@ build_dsdt(GArray *table_data, BIOSLinker *linker,aml_append (dev,aml_name_decl ("_CID" ,aml_eisaid ("PNP0A03" )));aml_append (dev,aml_name_decl ("_ADR" ,aml_int (0 )));aml_append (dev,aml_name_decl ("_UID" ,aml_int (1 ))); -aml_append (dev,aml_name_decl ("SUPP" ,aml_int (0 ))); -aml_append (dev,aml_name_decl ("CTRL" ,aml_int (0 )));aml_append (dev,build_q35_osc_method ());aml_append (sb_scope, dev);aml_append (dsdt, sb_scope);
Hi Michael,
Thanks, I'll use the above patch and send a new simplified version. Marcel