x86/hyperv: Fix the detection of E820_TYPE_PRAM in a Gen2 VM
Subsystems (3)
Dexuan Cui
maintainerPATCH 1/1 x86/hyperv: Fix the detection of E820_TYPE_PRAM in a Gen2 VM
A Gen2 VM doesn't support legacy PCI/PCIe, so both raw_pci_ops and raw_pci_ext_ops are NULL, and pci_subsys_init() -> pcibios_init() doesn't call pcibios_resource_survey() -> e820__reserve_resources_late(); as a result, any emulated persistent memory of E820_TYPE_PRAM (12) via the kernel parameter memmap=nn[KMG]!ss is not added into iomem_resource and hence can't be detected by register_e820_pmem().
Fix this by directly calling e820__reserve_resources_late() in hv_pci_init(), which is called from arch_initcall(pci_arch_init).
It's ok to move a Gen2 VM's e820__reserve_resources_late() from subsys_initcall(pci_subsys_init) to arch_initcall(pci_arch_init) because the code in-between doesn't depend on the E820 resources. e820__reserve_resources_late() depends on e820__reserve_resources(), which has been called earlier from setup_arch().
For a Gen-2 VM, the new hv_pci_init() also adds any memory of E820_TYPE_PMEM (7) into iomem_resource, and acpi_nfit_register_region() -> acpi_nfit_insert_resource() -> region_intersects() returns REGION_INTERSECTS, so the memory of E820_TYPE_PMEM won't get added twice.
Changed the local variable "int gen2vm" to "bool gen2vm".
--- arch/x86/hyperv/hv_init.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-)
arch/x86/hyperv/hv_init.c +21 -4
@@ -13,6 +13,7 @@#include <linux/io.h> #include <asm/apic.h> #include <asm/desc.h> +#include <asm/e820/api.h> #include <asm/sev.h> #include <asm/hypervisor.h> #include <asm/hyperv-tlfs.h>
@@ -282,15 +283,31 @@ static int hv_cpu_die(unsigned int cpu)static int __init hv_pci_init (void ) { -int gen2vm =efi_enabled (EFI_BOOT ); +bool gen2vm =efi_enabled (EFI_BOOT );/* -* For Generation-2 VM, we exit from pci_arch_init() by returning 0. -* The purpose is to suppress the harmless warning: +* A Generation-2 VM doesn't support legacy PCI/PCIe, so both +* raw_pci_ops and raw_pci_ext_ops are NULL, and pci_subsys_init() -> +* pcibios_init() doesn't call pcibios_resource_survey() -> +* e820__reserve_resources_late(); as a result, any emulated persistent +* memory of E820_TYPE_PRAM (12) via the kernel parameter +* memmap=nn[KMG]!ss is not added into iomem_resource and hence can't be +* detected by register_e820_pmem(). Fix this by directly calling +* e820__reserve_resources_late() here: e820__reserve_resources_late() +* depends on e820__reserve_resources(), which has been called earlier +* from setup_arch(). Note: e820__reserve_resources_late() also adds +* any memory of E820_TYPE_PMEM (7) into iomem_resource, and +* acpi_nfit_register_region() -> acpi_nfit_insert_resource() -> +* region_intersects() returns REGION_INTERSECTS, so the memory of +* E820_TYPE_PMEM won't get added twice. +* +* We return 0 here so that pci_arch_init() won't print the warning: * "PCI: Fatal: No config space access function found" */ -if (gen2vm) +if (gen2vm) { +e820__reserve_resources_late ();return 0 ; + }/* For Generation-1 VM, we'll proceed in pci_arch_init(). */ return 1 ;
Signature · 1 line
2.25.1
Quoted 64 lines · Dexuan Cui
A Gen2 VM doesn't support legacy PCI/PCIe, so both
raw_pci_opsandraw_pci_ext_opsare NULL, andpci_subsys_init()->pcibios_init()doesn't callpcibios_resource_survey()->e820__reserve_resources_late(); as a result, any emulated persistent memory ofE820_TYPE_PRAM(12) via the kernel parameter memmap=nn[KMG]!ss is not added intoiomem_resourceand hence can't be detected byregister_e820_pmem().Fix this by directly calling
e820__reserve_resources_late()inhv_pci_init(), which is called fromarch_initcall(pci_arch_init).It's ok to move a Gen2 VM's
e820__reserve_resources_late()fromsubsys_initcall(pci_subsys_init)toarch_initcall(pci_arch_init)because the code in-between doesn't depend on the E820 resources.e820__reserve_resources_late()depends one820__reserve_resources(), which has been called earlier fromsetup_arch().For a Gen-2 VM, the new
hv_pci_init()also adds any memory ofE820_TYPE_PMEM(7) intoiomem_resource, andacpi_nfit_register_region()->acpi_nfit_insert_resource()->region_intersects()returnsREGION_INTERSECTS, so the memory ofE820_TYPE_PMEMwon't get added twice.Changed the local variable "int gen2vm" to "bool gen2vm".
Signed-off-by Saurabh Sengar <[email protected]>Signed-off-by Dexuan Cui <[email protected]>--- arch/x86/hyperv/hv_init.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-)arch/x86/hyperv/hv_init.c +15
-3@@ -13,6 +13,7 @@#include <linux/io.h> #include <asm/apic.h> #include <asm/desc.h> +#include <asm/e820/api.h> #include <asm/sev.h> #include <asm/hypervisor.h> #include <asm/hyperv-tlfs.h> @@ -282,15 +283,31 @@ static int hv_cpu_die(unsigned int cpu)static int __init hv_pci_init (void ) { -int gen2vm =efi_enabled (EFI_BOOT ); +bool gen2vm =efi_enabled (EFI_BOOT );/* -* For Generation-2 VM, we exit from pci_arch_init() by returning 0. -* The purpose is to suppress the harmless warning: +* A Generation-2 VM doesn't support legacy PCI/PCIe, so both +* raw_pci_ops and raw_pci_ext_ops are NULL, and pci_subsys_init() -> +* pcibios_init() doesn't call pcibios_resource_survey() -> +* e820__reserve_resources_late(); as a result, any emulated persistent +* memory of E820_TYPE_PRAM (12) via the kernel parameter +* memmap=nn[KMG]!ss is not added into iomem_resource and hence can't be +* detected by register_e820_pmem(). Fix this by directly calling +* e820__reserve_resources_late() here: e820__reserve_resources_late() +* depends on e820__reserve_resources(), which has been called earlier +* from setup_arch(). Note: e820__reserve_resources_late() also adds +* any memory of E820_TYPE_PMEM (7) into iomem_resource, and +* acpi_nfit_register_region() -> acpi_nfit_insert_resource() -> +* region_intersects() returns REGION_INTERSECTS, so the memory of
arch/x86/hyperv/hv_init.c static int hv_cpu_die(unsigned int cpu) · line 310
Quoted 10 lines
arch/x86/hyperv/hv_init.c +6
-1+* E820_TYPE_PMEM won't get added twice. +* +* We return 0 here so that pci_arch_init() won't print the warning: * "PCI: Fatal: No config space access function found" */ -if (gen2vm) +if (gen2vm) { +e820__reserve_resources_late ();return 0 ; + }
Kind reminder to review this.
- Saurabh
Quoted 7 lines
arch/x86/hyperv/hv_init.c +0
-0/* For Generation-1 VM, we'll proceed in pci_arch_init(). */ return 1 ;Signature · 1 line
2.25.1
Wei Liu
maintainerQuoted 70 lines · Saurabh Singh Sengar
Quoted 69 lines · Dexuan Cui
A Gen2 VM doesn't support legacy PCI/PCIe, so both
raw_pci_opsandraw_pci_ext_opsare NULL, andpci_subsys_init()->pcibios_init()doesn't callpcibios_resource_survey()->e820__reserve_resources_late(); as a result, any emulated persistent memory ofE820_TYPE_PRAM(12) via the kernel parameter memmap=nn[KMG]!ss is not added intoiomem_resourceand hence can't be detected byregister_e820_pmem().Fix this by directly calling
e820__reserve_resources_late()inhv_pci_init(), which is called fromarch_initcall(pci_arch_init).It's ok to move a Gen2 VM's
e820__reserve_resources_late()fromsubsys_initcall(pci_subsys_init)toarch_initcall(pci_arch_init)because the code in-between doesn't depend on the E820 resources.e820__reserve_resources_late()depends one820__reserve_resources(), which has been called earlier fromsetup_arch().For a Gen-2 VM, the new
hv_pci_init()also adds any memory ofE820_TYPE_PMEM(7) intoiomem_resource, andacpi_nfit_register_region()->acpi_nfit_insert_resource()->region_intersects()returnsREGION_INTERSECTS, so the memory ofE820_TYPE_PMEMwon't get added twice.Changed the local variable "int gen2vm" to "bool gen2vm".
Signed-off-by Saurabh Sengar <[email protected]>Signed-off-by Dexuan Cui <[email protected]>--- arch/x86/hyperv/hv_init.c | 25 +++++++++++++++++++++---- 1 file changed, 21 insertions(+), 4 deletions(-)arch/x86/hyperv/hv_init.c +18
-3@@ -13,6 +13,7 @@#include <linux/io.h> #include <asm/apic.h> #include <asm/desc.h> +#include <asm/e820/api.h> #include <asm/sev.h> #include <asm/hypervisor.h> #include <asm/hyperv-tlfs.h> @@ -282,15 +283,31 @@ static int hv_cpu_die(unsigned int cpu)static int __init hv_pci_init (void ) { -int gen2vm =efi_enabled (EFI_BOOT ); +bool gen2vm =efi_enabled (EFI_BOOT );/* -* For Generation-2 VM, we exit from pci_arch_init() by returning 0. -* The purpose is to suppress the harmless warning: +* A Generation-2 VM doesn't support legacy PCI/PCIe, so both +* raw_pci_ops and raw_pci_ext_ops are NULL, and pci_subsys_init() -> +* pcibios_init() doesn't call pcibios_resource_survey() -> +* e820__reserve_resources_late(); as a result, any emulated persistent +* memory of E820_TYPE_PRAM (12) via the kernel parameter +* memmap=nn[KMG]!ss is not added into iomem_resource and hence can't be +* detected by register_e820_pmem(). Fix this by directly calling +* e820__reserve_resources_late() here: e820__reserve_resources_late() +* depends on e820__reserve_resources(), which has been called earlier +* from setup_arch(). Note: e820__reserve_resources_late() also adds +* any memory of E820_TYPE_PMEM (7) into iomem_resource, and +* acpi_nfit_register_region() -> acpi_nfit_insert_resource() -> +* region_intersects() returns REGION_INTERSECTS, so the memory of +* E820_TYPE_PMEM won't get added twice. +* +* We return 0 here so that pci_arch_init() won't print the warning: * "PCI: Fatal: No config space access function found" */
Quoted 9 lines
arch/x86/hyperv/hv_init.c static int hv_cpu_die(unsigned int cpu) · line 310
Quoted 5 lines
arch/x86/hyperv/hv_init.c +3
-1-if (gen2vm) +if (gen2vm) { +e820__reserve_resources_late ();return 0 ; + }Kind reminder to review this.
I tried to applied this patch to hyperv-fixes, but it doesn't apply cleanly.
Please resend.
Thanks,
Wei.