VMD endpoint passthrough support
Subsystems (4)
v2 cover VMD endpoint passthrough support
This set contains 2 patches for Linux and 1 for QEMU. VMD device 8086:28C0 contains information in registers to assist with direct assignment passthrough. Several other VMD devices don't have this information, but can easily be emulated to offer this feature.
The existing VMD devices not supporting the feature cannot be changed to offer the information, but also don't restrict the ability to offer this information in emulation by the hypervisor. Future VMD devices will offer the 28C0 mode natively.
The QEMU patch emulates the hardware assistance that the VMD 28C0 device
provides: a config space register claiming passthrough support, and the shadow membar registers containing the host information for guest address assignment in the VMD domain. These VMD devices have this config space register set as reserved and will not conflict with the emulated bit.
The Linux patch allows guest kernels to use the passthrough information emulated by the QEMU patch, by matching the config space register claiming passthrough support.
Changes from v1:
v1 changed the VMD Subsystem ID to QEMU's so that the guest driver could match against it. This was unnecessary as the VMLOCK register and shadow membar registers could be safely emulated. Future VMDs will be aligned on these register bits.
Added the resource bit filtering patch that got lost in the mailserver.
v1: https://lore.kernel.org/linux-pci/[email protected]/
Jon Derrick (2):
PCI: vmd: Filter resource type bits from shadow register
PCI: vmd: Use Shadow MEMBAR registers for QEMU/KVM guests
drivers/pci/controller/vmd.c | 21 ++++++++++++++------- 1 file changed, 14 insertions(+), 7 deletions(-)
Signature · 1 line
2.18.1
PATCH v2 1/1 hw/vfio: Add VMD Passthrough Quirk
The VMD endpoint provides a real PCIe domain to the guest, including bridges and endpoints. Because the VMD domain is enumerated by the guest kernel, the guest kernel will assign Guest Physical Addresses to the downstream endpoint BARs and bridge windows.
When the guest kernel performs MMIO to VMD sub-devices, IOMMU will translate from the guest address space to the physical address space. Because the bridges have been programmed with guest addresses, the bridges will reject the transaction containing physical addresses.
VMD device 28C0 natively assists passthrough by providing the Host Physical Address in shadow registers accessible to the guest for bridge window assignment. The shadow registers are valid if bit 1 is set in VMD VMLOCK config register 0x70. Future VMDs will also support this feature. Existing VMDs have config register 0x70 reserved, and will return 0 on reads.
In order to support existing VMDs, this quirk emulates the VMLOCK and HPA shadow registers for all VMD device ids which don't natively assist with passthrough. The Linux VMD driver is updated to allow existing VMD devices to query VMLOCK for passthrough support.
--- hw/vfio/pci-quirks.c | 103 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/pci.c | 7 +++ hw/vfio/pci.h | 2 + hw/vfio/trace-events | 3 ++ 4 files changed, 115 insertions(+)
hw/vfio/pci-quirks.c +103 -0
@@ -1709,3 +1709,106 @@ free_exit:return ret; } + +/* +* The VMD endpoint provides a real PCIe domain to the guest and the guest +* kernel performs enumeration of the VMD sub-device domain. Guest transactions +* to VMD sub-devices go through IOMMU translation from guest addresses to +* physical addresses. When MMIO goes to an endpoint after being translated to +* physical addresses, the bridge rejects the transaction because the window +* has been programmed with guest addresses. +* +* VMD can use the Host Physical Address in order to correctly program the +* bridge windows in its PCIe domain. VMD device 28C0 has HPA shadow registers +* located at offset 0x2000 in MEMBAR2 (BAR 4). The shadow registers are valid +* if bit 1 is set in the VMD VMLOCK config register 0x70. VMD devices without +* this native assistance can have these registers safely emulated as these +* registers are reserved. +*/ +typedef struct VFIOVMDQuirk { +VFIOPCIDevice *vdev; +uint64_t membar_phys[2 ]; +}VFIOVMDQuirk ; + +static uint64_t vfio_vmd_quirk_read (void *opaque,hwaddr addr,unsigned size) +{ +VFIOVMDQuirk *data = opaque; +uint64_t val =0 ; + +memcpy (&val, (void *)data->membar_phys + addr, size); +return val; +} + +static const MemoryRegionOps vfio_vmd_quirk = { + .read = vfio_vmd_quirk_read, + .endianness =DEVICE_LITTLE_ENDIAN , +}; + +#define VMD_VMLOCK 0x70 +#define VMD_SHADOW 0x2000 +#define VMD_MEMBAR2 4 + +static int vfio_vmd_emulate_shadow_registers (VFIOPCIDevice *vdev) +{ +VFIOQuirk *quirk; +VFIOVMDQuirk *data; +PCIDevice *pdev = &vdev->pdev; +int ret; + + data =g_malloc0 (sizeof (*data)); + ret =pread (vdev->vbasedev.fd, data->membar_phys,16 , + vdev->config_offset +PCI_BASE_ADDRESS_2 ); +if (ret !=16 ) { +error_report ("VMD %s cannot read MEMBARs (%d)" , + vdev->vbasedev.name, ret); +g_free (data); +return -EFAULT ; + } + + quirk =vfio_quirk_alloc (1 ); + quirk->data = data; + data->vdev = vdev; + +/* Emulate Shadow Registers */ +memory_region_init_io (quirk->mem,OBJECT (vdev), &vfio_vmd_quirk, data, +"vfio-vmd-quirk" ,sizeof (data->membar_phys)); +memory_region_add_subregion_overlap (vdev->bars[VMD_MEMBAR2 ].region.mem, +VMD_SHADOW , quirk->mem,1 ); +memory_region_set_readonly (quirk->mem, true); +memory_region_set_enabled (quirk->mem, true); + +QLIST_INSERT_HEAD (&vdev->bars[VMD_MEMBAR2 ].quirks, quirk, next); + +trace_vfio_pci_vmd_quirk_shadow_regs (vdev->vbasedev.name, + data->membar_phys[0 ], + data->membar_phys[1 ]); + +/* Advertise Shadow Register support */ +pci_byte_test_and_set_mask (pdev->config +VMD_VMLOCK ,0x2 ); +pci_set_byte (pdev->wmask +VMD_VMLOCK ,0 ); +pci_set_byte (vdev->emulated_config_bits +VMD_VMLOCK ,0x2 ); + +trace_vfio_pci_vmd_quirk_vmlock (vdev->vbasedev.name, +pci_get_byte (pdev->config +VMD_VMLOCK )); + +return 0 ; +} + +int vfio_pci_vmd_init (VFIOPCIDevice *vdev) +{ +int ret =0 ; + +switch (vdev->device_id) { +case 0x28C0 :/* Native passthrough support */ +break ; +/* Emulates Native passthrough support */ +case 0x201D : +case 0x467F : +case 0x4C3D : +case 0x9A0B : + ret =vfio_vmd_emulate_shadow_registers (vdev); +break ; + } + +return ret; +}
hw/vfio/pci.c +7 -0
@@ -3024,6 +3024,13 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) } } +if (vdev->vendor_id ==PCI_VENDOR_ID_INTEL ) { + ret =vfio_pci_vmd_init (vdev); +if (ret) { +error_report ("Failed to setup VMD" ); + } + } +vfio_register_err_notifier (vdev);vfio_register_req_notifier (vdev);vfio_setup_resetfn_quirk (vdev);
hw/vfio/pci.h +2 -0
@@ -217,6 +217,8 @@ int vfio_pci_igd_opregion_init(VFIOPCIDevice *vdev,int vfio_pci_nvidia_v100_ram_init (VFIOPCIDevice *vdev,Error **errp);int vfio_pci_nvlink2_init (VFIOPCIDevice *vdev,Error **errp); +int vfio_pci_vmd_init (VFIOPCIDevice *vdev); +void vfio_display_reset (VFIOPCIDevice *vdev);int vfio_display_probe (VFIOPCIDevice *vdev,Error **errp);void vfio_display_finalize (VFIOPCIDevice *vdev);
hw/vfio/trace-events +3 -0
@@ -90,6 +90,9 @@ vfio_pci_nvidia_gpu_setup_quirk(const char *name, uint64_t tgt, uint64_t size) " vfio_pci_nvlink2_setup_quirk_ssatgt(const char *name, uint64_t tgt, uint64_t size) "%s tgt=0x%"PRIx64" size=0x%"PRIx64 vfio_pci_nvlink2_setup_quirk_lnkspd(const char *name, uint32_t link_speed) "%s link_speed=0x%x" +vfio_pci_vmd_quirk_shadow_regs(const char *name, uint64_t mb1, uint64_t mb2) "%s membar1_phys=0x%"PRIx64" membar2_phys=0x%"PRIx64 +vfio_pci_vmd_quirk_vmlock(const char *name, uint8_t vmlock) "%s vmlock=0x%x" + # common.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64
Signature · 1 line
2.18.1
Re: [PATCH for QEMU v2] hw/vfio: Add VMD Passthrough Quirk
Quoted 1 line · On Mon, 11 May 2020 15:01:27 -0400 Jon Derrick <[email protected]>…
The VMD endpoint provides a real PCIe domain to the guest, including
Please define VMD. I'm sure this is obvious to many, but I've had to do some research. The best TL;DR summary I've found is Keith's original commit 185a383ada2e adding the controller to Linux. If there's something better, please let me know.
Quoted 8 lines
bridges and endpoints. Because the VMD domain is enumerated by the guest kernel, the guest kernel will assign Guest Physical Addresses to the downstream endpoint BARs and bridge windows.
When the guest kernel performs MMIO to VMD sub-devices, IOMMU will translate from the guest address space to the physical address space. Because the bridges have been programmed with guest addresses, the bridges will reject the transaction containing physical addresses.
I'm lost, what IOMMU is involved in CPU access to MMIO space? My guess is that since all MMIO of this domain is mapped behind the host endpoint BARs 2 & 4 that QEMU simply accesses it via mapping of those BARs into the VM, so it's the MMU, not the IOMMU performing those GPA to HPA translations. But then presumably the bridges within the domain are scrambled because their apertures are programmed with ranges that don't map into the VMD endpoint BARs. Is that remotely correct? Some /proc/iomem output and/or lspci listing from the host to see how this works would be useful.
Quoted 6 lines
VMD device 28C0 natively assists passthrough by providing the Host Physical Address in shadow registers accessible to the guest for bridge window assignment. The shadow registers are valid if bit 1 is set in VMD VMLOCK config register
0x70. Future VMDs will also support this feature. Existing VMDs have config register0x70reserved, and will return 0 on reads.
So these shadow registers are simply exposing the host BAR2 & BAR4 addresses into the guest, so the quirk is dependent on reading those values from the device before anyone has written to them and the BAR emulation in the kernel kicks in (not a problem, just an observation).
Does the VMD controller code then use these bases addresses to program the bridges/endpoint within the domain? What does the same /proc/iomem or lspci look like inside the guest then? It seems like we'd see the VMD endpoint with GPA BARs, but the devices within the domain using HPAs. If that's remotely true, and we're not forcing an identity mapping of this HPA range into the GPA, does the vmd controller driver impose a TRA function on these MMIO addresses in the guest?
Sorry if I'm way off, I'm piecing things together from scant information here. Please Cc me on future vfio related patches. Thanks,
Alex
Quoted 169 lines
In order to support existing VMDs, this quirk emulates the VMLOCK and HPA shadow registers for all VMD device ids which don't natively assist with passthrough. The Linux VMD driver is updated to allow existing VMD devices to query VMLOCK for passthrough support.
Signed-off-by Jon Derrick <[email protected]>--- hw/vfio/pci-quirks.c | 103 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/pci.c | 7 +++ hw/vfio/pci.h | 2 + hw/vfio/trace-events | 3 ++ 4 files changed, 115 insertions(+)hw/vfio/pci-quirks.c +103
-0@@ -1709,3 +1709,106 @@ free_exit:return ret; } + +/* +* The VMD endpoint provides a real PCIe domain to the guest and the guest +* kernel performs enumeration of the VMD sub-device domain. Guest transactions +* to VMD sub-devices go through IOMMU translation from guest addresses to +* physical addresses. When MMIO goes to an endpoint after being translated to +* physical addresses, the bridge rejects the transaction because the window +* has been programmed with guest addresses. +* +* VMD can use the Host Physical Address in order to correctly program the +* bridge windows in its PCIe domain. VMD device 28C0 has HPA shadow registers +* located at offset 0x2000 in MEMBAR2 (BAR 4). The shadow registers are valid +* if bit 1 is set in the VMD VMLOCK config register 0x70. VMD devices without +* this native assistance can have these registers safely emulated as these +* registers are reserved. +*/ +typedef struct VFIOVMDQuirk { +VFIOPCIDevice *vdev; +uint64_t membar_phys[2 ]; +}VFIOVMDQuirk ; + +static uint64_t vfio_vmd_quirk_read (void *opaque,hwaddr addr,unsigned size) +{ +VFIOVMDQuirk *data = opaque; +uint64_t val =0 ; + +memcpy (&val, (void *)data->membar_phys + addr, size); +return val; +} + +static const MemoryRegionOps vfio_vmd_quirk = { + .read = vfio_vmd_quirk_read, + .endianness =DEVICE_LITTLE_ENDIAN , +}; + +#define VMD_VMLOCK 0x70 +#define VMD_SHADOW 0x2000 +#define VMD_MEMBAR2 4 + +static int vfio_vmd_emulate_shadow_registers (VFIOPCIDevice *vdev) +{ +VFIOQuirk *quirk; +VFIOVMDQuirk *data; +PCIDevice *pdev = &vdev->pdev; +int ret; + + data =g_malloc0 (sizeof (*data)); + ret =pread (vdev->vbasedev.fd, data->membar_phys,16 , + vdev->config_offset +PCI_BASE_ADDRESS_2 ); +if (ret !=16 ) { +error_report ("VMD %s cannot read MEMBARs (%d)" , + vdev->vbasedev.name, ret); +g_free (data); +return -EFAULT ; + } + + quirk =vfio_quirk_alloc (1 ); + quirk->data = data; + data->vdev = vdev; + +/* Emulate Shadow Registers */ +memory_region_init_io (quirk->mem,OBJECT (vdev), &vfio_vmd_quirk, data, +"vfio-vmd-quirk" ,sizeof (data->membar_phys)); +memory_region_add_subregion_overlap (vdev->bars[VMD_MEMBAR2 ].region.mem, +VMD_SHADOW , quirk->mem,1 ); +memory_region_set_readonly (quirk->mem, true); +memory_region_set_enabled (quirk->mem, true); + +QLIST_INSERT_HEAD (&vdev->bars[VMD_MEMBAR2 ].quirks, quirk, next); + +trace_vfio_pci_vmd_quirk_shadow_regs (vdev->vbasedev.name, + data->membar_phys[0 ], + data->membar_phys[1 ]); + +/* Advertise Shadow Register support */ +pci_byte_test_and_set_mask (pdev->config +VMD_VMLOCK ,0x2 ); +pci_set_byte (pdev->wmask +VMD_VMLOCK ,0 ); +pci_set_byte (vdev->emulated_config_bits +VMD_VMLOCK ,0x2 ); + +trace_vfio_pci_vmd_quirk_vmlock (vdev->vbasedev.name, +pci_get_byte (pdev->config +VMD_VMLOCK )); + +return 0 ; +} + +int vfio_pci_vmd_init (VFIOPCIDevice *vdev) +{ +int ret =0 ; + +switch (vdev->device_id) { +case 0x28C0 :/* Native passthrough support */ +break ; +/* Emulates Native passthrough support */ +case 0x201D : +case 0x467F : +case 0x4C3D : +case 0x9A0B : + ret =vfio_vmd_emulate_shadow_registers (vdev); +break ; + } + +return ret; +}hw/vfio/pci.c +7
-0@@ -3024,6 +3024,13 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) } } +if (vdev->vendor_id ==PCI_VENDOR_ID_INTEL ) { + ret =vfio_pci_vmd_init (vdev); +if (ret) { +error_report ("Failed to setup VMD" ); + } + } +vfio_register_err_notifier (vdev);vfio_register_req_notifier (vdev);vfio_setup_resetfn_quirk (vdev);hw/vfio/pci.h +2
-0@@ -217,6 +217,8 @@ int vfio_pci_igd_opregion_init(VFIOPCIDevice *vdev,int vfio_pci_nvidia_v100_ram_init (VFIOPCIDevice *vdev,Error **errp);int vfio_pci_nvlink2_init (VFIOPCIDevice *vdev,Error **errp); +int vfio_pci_vmd_init (VFIOPCIDevice *vdev); +void vfio_display_reset (VFIOPCIDevice *vdev);int vfio_display_probe (VFIOPCIDevice *vdev,Error **errp);void vfio_display_finalize (VFIOPCIDevice *vdev);hw/vfio/trace-events +3
-0@@ -90,6 +90,9 @@ vfio_pci_nvidia_gpu_setup_quirk(const char *name, uint64_t tgt, uint64_t size) " vfio_pci_nvlink2_setup_quirk_ssatgt(const char *name, uint64_t tgt, uint64_t size) "%s tgt=0x%"PRIx64" size=0x%"PRIx64 vfio_pci_nvlink2_setup_quirk_lnkspd(const char *name, uint32_t link_speed) "%s link_speed=0x%x" +vfio_pci_vmd_quirk_shadow_regs(const char *name, uint64_t mb1, uint64_t mb2) "%s membar1_phys=0x%"PRIx64" membar2_phys=0x%"PRIx64 +vfio_pci_vmd_quirk_vmlock(const char *name, uint8_t vmlock) "%s vmlock=0x%x" + # common.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64
Re: [PATCH for QEMU v2] hw/vfio: Add VMD Passthrough Quirk
Hi Alex,
I'm probably not getting the translation technical details correct.
Quoted 9 lines · Alex Williamson
Quoted 1 line · On Mon, 11 May 2020 15:01:27 -0400 Jon Derrick <[email protected]>…
The VMD endpoint provides a real PCIe domain to the guest, including
Please define VMD. I'm sure this is obvious to many, but I've had to do some research. The best TL;DR summary I've found is Keith's original commit
185a383ada2eadding the controller to Linux. If there's something better, please let me know.
That's the correct commit, but I'll try to summarize the important bits for v3.
Quoted 10 lines
Quoted 9 lines
bridges and endpoints. Because the VMD domain is enumerated by the guest kernel, the guest kernel will assign Guest Physical Addresses to the downstream endpoint BARs and bridge windows.
When the guest kernel performs MMIO to VMD sub-devices, IOMMU will translate from the guest address space to the physical address space. Because the bridges have been programmed with guest addresses, the bridges will reject the transaction containing physical addresses.
Quoted 9 lines
I'm lost, what IOMMU is involved in CPU access to MMIO space? My guess is that since all MMIO of this domain is mapped behind the host endpoint BARs 2 & 4 that QEMU simply accesses it via mapping of those BARs into the VM, so it's the MMU, not the IOMMU performing those GPA to HPA translations. But then presumably the bridges within the domain are scrambled because their apertures are programmed with ranges that don't map into the VMD endpoint BARs. Is that remotely correct? Some
/proc/iomemoutput and/or lspci listing from the host to see how this works would be useful.
Correct. So MMU not IOMMU.
In the guest kernel, the bridges and devices in the VMD domain are programmed with the addresses provided in the VMD endpoint's BAR2&4 (MEMBAR1&2). Because these BARs are populated with guest addresses, MMU translates to host physical and the bridge window rejects MMIO not in its [GPA] range.
As an example:
Host:
94000000-97ffffff : 0000:17:05.5
94000000-97ffffff : VMD MEMBAR1
94000000-943fffff : PCI Bus 10000:01
94000000-9400ffff : 10000:01:00.0
94010000-94013fff : 10000:01:00.0
94010000-94013fff : nvme
94400000-947fffff : PCI Bus 10000:01
94800000-94bfffff : PCI Bus 10000:02
94800000-9480ffff : 10000:02:00.0
94810000-94813fff : 10000:02:00.0
94810000-94813fff : nvme
94c00000-94ffffff : PCI Bus 10000:02MEMBAR 2 is similarly assigned
Quoted 8 lines
Quoted 7 lines
VMD device 28C0 natively assists passthrough by providing the Host Physical Address in shadow registers accessible to the guest for bridge window assignment. The shadow registers are valid if bit 1 is set in VMD VMLOCK config register
0x70. Future VMDs will also support this feature. Existing VMDs have config register0x70reserved, and will return 0 on reads.
Quoted 4 lines
So these shadow registers are simply exposing the host BAR2 & BAR4 addresses into the guest, so the quirk is dependent on reading those values from the device before anyone has written to them and the BAR emulation in the kernel kicks in (not a problem, just an observation).
It's not expected that there will be anything writing that resource and those registers are read-only.
The first 0x2000 of MEMBAR2 (BAR4) contain msix tables, and mappings to subordinate buses are on 1MB aligned.
Quoted 7 lines
Does the VMD controller code then use these bases addresses to program the bridges/endpoint within the domain? What does the same
/proc/iomemor lspci look like inside the guest then? It seems like we'd see the VMD endpoint with GPA BARs, but the devices within the domain using HPAs. If that's remotely true, and we're not forcing an identity mapping of this HPA range into the GPA, does the vmd controller driver impose a TRA function on these MMIO addresses in the guest?
This is the guest with the guest addresses:
f8000000-fbffffff : 0000:00:07.0
f8000000-fbffffff : VMD MEMBAR1
f8000000-f83fffff : PCI Bus 10000:01
f8000000-f800ffff :
10000:01:00.0
f8010000-f8013fff : 10000:01:00.0
f801000
0-f8013fff : nvme
f8400000-f87fffff : PCI Bus 10000:01
f88000
00-f8bfffff : PCI Bus 10000:02
f8800000-f880ffff : 10000:02:00.0
f8810000-f8813fff : 10000:02:00.0
f8810000-f8813fff :
nvme
f8c00000-f8ffffff : PCI Bus 10000:02The VMD guest driver does the translation on the supplied address using pci_add_resource_offset prior to creating the root bus and enumerating the domain:
offset[0] = vmd->dev->resource[VMD_MEMBAR1].start - readq(membar2 + MB2_SHADOW_OFFSET); offset[1] = vmd->dev->resource[VMD_MEMBAR2].start - readq(membar2 + MB2_SHADOW_OFFSET + 8); ... pci_add_resource(&resources, &vmd->resources[0]); pci_add_resource_offset(&resources, &vmd->resources[1], offset[0]); pci_add_resource_offset(&resources, &vmd->resources[2], offset[1]); vmd->bus = pci_create_root_bus(&vmd->dev->dev, vmd->busn_start, &vmd_ops, sd, &resources);
The offset becomes the CPU-to-bridge translation function for programming the guest's VMD domain with bus addresses.
In the patched guest's lspci, for the bridge you see:
# lspci -v -s 10000:00:02.0 10000:00:02.0 PCI bridge: Intel Corporation Sky Lake-E PCI Express Root Port C ... Memory behind bridge: 94000000-943fffff
But the kernel doesn't export the offset BAR for the endpoint:
# lspci -v -s 10000:02:00.0 10000:02:00.0 Non-Volatile memory controller: Intel Corporation NVMe Datacenter SSD ... Memory at f8810000 (64-bit, non-prefetchable) [size=16K] [virtual] Expansion ROM at f8800000 [disabled] [size=64K]
The endpoint is still programmed with the offset:
# setpci -s 10000:02:00.0 10.l 94810004
Quoted 6 lines
Sorry if I'm way off, I'm piecing things together from scant information here. Please Cc me on future vfio related patches. Thanks,
Alex
No problem
Thanks,
Jon
Quoted 170 lines
Quoted 170 lines
In order to support existing VMDs, this quirk emulates the VMLOCK and HPA shadow registers for all VMD device ids which don't natively assist with passthrough. The Linux VMD driver is updated to allow existing VMD devices to query VMLOCK for passthrough support.
Signed-off-by Jon Derrick <[email protected]>--- hw/vfio/pci-quirks.c | 103 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/pci.c | 7 +++ hw/vfio/pci.h | 2 + hw/vfio/trace-events | 3 ++ 4 files changed, 115 insertions(+)hw/vfio/pci-quirks.c +103
-0@@ -1709,3 +1709,106 @@ free_exit:return ret; } + +/* +* The VMD endpoint provides a real PCIe domain to the guest and the guest +* kernel performs enumeration of the VMD sub-device domain. Guest transactions +* to VMD sub-devices go through IOMMU translation from guest addresses to +* physical addresses. When MMIO goes to an endpoint after being translated to +* physical addresses, the bridge rejects the transaction because the window +* has been programmed with guest addresses. +* +* VMD can use the Host Physical Address in order to correctly program the +* bridge windows in its PCIe domain. VMD device 28C0 has HPA shadow registers +* located at offset 0x2000 in MEMBAR2 (BAR 4). The shadow registers are valid +* if bit 1 is set in the VMD VMLOCK config register 0x70. VMD devices without +* this native assistance can have these registers safely emulated as these +* registers are reserved. +*/ +typedef struct VFIOVMDQuirk { +VFIOPCIDevice *vdev; +uint64_t membar_phys[2 ]; +}VFIOVMDQuirk ; + +static uint64_t vfio_vmd_quirk_read (void *opaque,hwaddr addr,unsigned size) +{ +VFIOVMDQuirk *data = opaque; +uint64_t val =0 ; + +memcpy (&val, (void *)data->membar_phys + addr, size); +return val; +} + +static const MemoryRegionOps vfio_vmd_quirk = { + .read = vfio_vmd_quirk_read, + .endianness =DEVICE_LITTLE_ENDIAN , +}; + +#define VMD_VMLOCK 0x70 +#define VMD_SHADOW 0x2000 +#define VMD_MEMBAR2 4 + +static int vfio_vmd_emulate_shadow_registers (VFIOPCIDevice *vdev) +{ +VFIOQuirk *quirk; +VFIOVMDQuirk *data; +PCIDevice *pdev = &vdev->pdev; +int ret; + + data =g_malloc0 (sizeof (*data)); + ret =pread (vdev->vbasedev.fd, data->membar_phys,16 , + vdev->config_offset +PCI_BASE_ADDRESS_2 ); +if (ret !=16 ) { +error_report ("VMD %s cannot read MEMBARs (%d)" , + vdev->vbasedev.name, ret); +g_free (data); +return -EFAULT ; + } + + quirk =vfio_quirk_alloc (1 ); + quirk->data = data; + data->vdev = vdev; + +/* Emulate Shadow Registers */ +memory_region_init_io (quirk->mem,OBJECT (vdev), &vfio_vmd_quirk, data, +"vfio-vmd-quirk" ,sizeof (data->membar_phys)); +memory_region_add_subregion_overlap (vdev->bars[VMD_MEMBAR2 ].region.mem, +VMD_SHADOW , quirk->mem,1 ); +memory_region_set_readonly (quirk->mem, true); +memory_region_set_enabled (quirk->mem, true); + +QLIST_INSERT_HEAD (&vdev->bars[VMD_MEMBAR2 ].quirks, quirk, next); + +trace_vfio_pci_vmd_quirk_shadow_regs (vdev->vbasedev.name, + data->membar_phys[0 ], + data->membar_phys[1 ]); + +/* Advertise Shadow Register support */ +pci_byte_test_and_set_mask (pdev->config +VMD_VMLOCK ,0x2 ); +pci_set_byte (pdev->wmask +VMD_VMLOCK ,0 ); +pci_set_byte (vdev->emulated_config_bits +VMD_VMLOCK ,0x2 ); + +trace_vfio_pci_vmd_quirk_vmlock (vdev->vbasedev.name, +pci_get_byte (pdev->config +VMD_VMLOCK )); + +return 0 ; +} + +int vfio_pci_vmd_init (VFIOPCIDevice *vdev) +{ +int ret =0 ; + +switch (vdev->device_id) { +case 0x28C0 :/* Native passthrough support */ +break ; +/* Emulates Native passthrough support */ +case 0x201D : +case 0x467F : +case 0x4C3D : +case 0x9A0B : + ret =vfio_vmd_emulate_shadow_registers (vdev); +break ; + } + +return ret; +}hw/vfio/pci.c +7
-0@@ -3024,6 +3024,13 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) } } +if (vdev->vendor_id ==PCI_VENDOR_ID_INTEL ) { + ret =vfio_pci_vmd_init (vdev); +if (ret) { +error_report ("Failed to setup VMD" ); + } + } +vfio_register_err_notifier (vdev);vfio_register_req_notifier (vdev);vfio_setup_resetfn_quirk (vdev);hw/vfio/pci.h +2
-0@@ -217,6 +217,8 @@ int vfio_pci_igd_opregion_init(VFIOPCIDevice *vdev,int vfio_pci_nvidia_v100_ram_init (VFIOPCIDevice *vdev,Error **errp);int vfio_pci_nvlink2_init (VFIOPCIDevice *vdev,Error **errp); +int vfio_pci_vmd_init (VFIOPCIDevice *vdev); +void vfio_display_reset (VFIOPCIDevice *vdev);int vfio_display_probe (VFIOPCIDevice *vdev,Error **errp);void vfio_display_finalize (VFIOPCIDevice *vdev);hw/vfio/trace-events +3
-0@@ -90,6 +90,9 @@ vfio_pci_nvidia_gpu_setup_quirk(const char *name, uint64_t tgt, uint64_t size) " vfio_pci_nvlink2_setup_quirk_ssatgt(const char *name, uint64_t tgt, uint64_t size) "%s tgt=0x%"PRIx64" size=0x%"PRIx64 vfio_pci_nvlink2_setup_quirk_lnkspd(const char *name, uint32_t link_speed) "%s link_speed=0x%x" +vfio_pci_vmd_quirk_shadow_regs(const char *name, uint64_t mb1, uint64_t mb2) "%s membar1_phys=0x%"PRIx64" membar2_phys=0x%"PRIx64 +vfio_pci_vmd_quirk_vmlock(const char *name, uint8_t vmlock) "%s vmlock=0x%x" + # common.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64
Re: [PATCH for QEMU v2] hw/vfio: Add VMD Passthrough Quirk
Quoted 149 lines · Jonathan
Hi Alex,
I'm probably not getting the translation technical details correct.
Quoted 9 lines · Alex Williamson
Quoted 1 line · On Mon, 11 May 2020 15:01:27 -0400 Jon Derrick <[email protected]>…
The VMD endpoint provides a real PCIe domain to the guest, including
Please define VMD. I'm sure this is obvious to many, but I've had to do some research. The best TL;DR summary I've found is Keith's original commit
185a383ada2eadding the controller to Linux. If there's something better, please let me know.That's the correct commit, but I'll try to summarize the important bits for v3.
Quoted 10 lines
Quoted 9 lines
bridges and endpoints. Because the VMD domain is enumerated by the guest kernel, the guest kernel will assign Guest Physical Addresses to the downstream endpoint BARs and bridge windows.
When the guest kernel performs MMIO to VMD sub-devices, IOMMU will translate from the guest address space to the physical address space. Because the bridges have been programmed with guest addresses, the bridges will reject the transaction containing physical addresses.
Quoted 9 lines
I'm lost, what IOMMU is involved in CPU access to MMIO space? My guess is that since all MMIO of this domain is mapped behind the host endpoint BARs 2 & 4 that QEMU simply accesses it via mapping of those BARs into the VM, so it's the MMU, not the IOMMU performing those GPA to HPA translations. But then presumably the bridges within the domain are scrambled because their apertures are programmed with ranges that don't map into the VMD endpoint BARs. Is that remotely correct? Some
/proc/iomemoutput and/or lspci listing from the host to see how this works would be useful.Correct. So MMU not IOMMU.
In the guest kernel, the bridges and devices in the VMD domain are programmed with the addresses provided in the VMD endpoint's BAR2&4 (MEMBAR1&2). Because these BARs are populated with guest addresses, MMU translates to host physical and the bridge window rejects MMIO not in its [GPA] range.
As an example:
Host:
94000000-97ffffff : 0000:17:05.594000000-97ffffff : VMD MEMBAR1 94000000-943fffff : PCI Bus 10000:01 94000000-9400ffff : 10000:01:00.0 94010000-94013fff : 10000:01:00.0 94010000-94013fff : nvme 94400000-947fffff : PCI Bus 10000:01 94800000-94bfffff : PCI Bus 10000:02 94800000-9480ffff : 10000:02:00.0 94810000-94813fff : 10000:02:00.0 94810000-94813fff : nvme 94c00000-94ffffff : PCI Bus 10000:02MEMBAR 2 is similarly assigned
Quoted 8 lines
Quoted 7 lines
VMD device 28C0 natively assists passthrough by providing the Host Physical Address in shadow registers accessible to the guest for bridge window assignment. The shadow registers are valid if bit 1 is set in VMD VMLOCK config register
0x70. Future VMDs will also support this feature. Existing VMDs have config register0x70reserved, and will return 0 on reads.Quoted 4 lines
So these shadow registers are simply exposing the host BAR2 & BAR4 addresses into the guest, so the quirk is dependent on reading those values from the device before anyone has written to them and the BAR emulation in the kernel kicks in (not a problem, just an observation).
It's not expected that there will be anything writing that resource and those registers are read-only.
The first0x2000of MEMBAR2 (BAR4) contain msix tables, and mappings to subordinate buses are on 1MB aligned.Quoted 7 lines
Does the VMD controller code then use these bases addresses to program the bridges/endpoint within the domain? What does the same
/proc/iomemor lspci look like inside the guest then? It seems like we'd see the VMD endpoint with GPA BARs, but the devices within the domain using HPAs. If that's remotely true, and we're not forcing an identity mapping of this HPA range into the GPA, does the vmd controller driver impose a TRA function on these MMIO addresses in the guest?This is the guest with the guest addresses:
f8000000-fbffffff : 0000:00:07.0f8000000-fbffffff : VMD MEMBAR1 f8000000-f83fffff : PCI Bus 10000:01 f8000000-f800ffff : 10000:01:00.0 f8010000-f8013fff : 10000:01:00.0 f801000 0-f8013fff : nvme f8400000-f87fffff : PCI Bus 10000:01 f88000 00-f8bfffff : PCI Bus 10000:02 f8800000-f880ffff : 10000:02:00.0 f8810000-f8813fff : 10000:02:00.0 f8810000-f8813fff : nvme f8c00000-f8ffffff : PCI Bus 10000:02The VMD guest driver does the translation on the supplied address using
pci_add_resource_offsetprior to creating the root bus and enumerating the domain:offset[0] = vmd->dev->resource[VMD_MEMBAR1].start - readq(membar2 + MB2_SHADOW_OFFSET); offset[1] = vmd->dev->resource[VMD_MEMBAR2].start - readq(membar2 + MB2_SHADOW_OFFSET + 8); ... pci_add_resource(&resources, &vmd->resources[0]); pci_add_resource_offset(&resources, &vmd->resources[1], offset[0]); pci_add_resource_offset(&resources, &vmd->resources[2], offset[1]); vmd->bus = pci_create_root_bus(&vmd->dev->dev, vmd->busn_start, &vmd_ops, sd, &resources);The offset becomes the CPU-to-bridge translation function for programming the guest's VMD domain with bus addresses.
In the patched guest's lspci, for the bridge you see:
# lspci -v -s 10000:00:02.0 10000:00:02.0 PCI bridge: Intel Corporation Sky Lake-E PCI Express Root Port C ... Memory behind bridge: 94000000-943fffffBut the kernel doesn't export the offset BAR for the endpoint:
# lspci -v -s 10000:02:00.0 10000:02:00.0 Non-Volatile memory controller: Intel Corporation NVMe Datacenter SSD ... Memory at f8810000 (64-bit, non-prefetchable) [size=16K] [virtual] Expansion ROM at f8800000 [disabled] [size=64K]The endpoint is still programmed with the offset:
# setpci -s 10000:02:00.0 10.l 94810004Quoted 1 line
Quoted 6 lines
Quoted 5 lines
Sorry if I'm way off, I'm piecing things together from scant information here. Please Cc me on future vfio related patches. Thanks,
Alex
No problem
Thanks for the confirmation. The approach seems ok, but a subtle side-effect of MemoryRegion quirks is that we're going to trap the entire PAGE_SIZE range overlapping the quirk out to QEMU on guest access. The two registers at 0x2000 might be reserved for this purpose, but is there anything else that could be performance sensitive anywhere in that page? If so, it might be less transparent, but more efficient to invent a new quirk making use of config space (ex. adding an emulated vendor capability somewhere known to be unused on the device). Thanks,
Alex
Quoted 169 lines
Quoted 169 lines
Quoted 169 lines
In order to support existing VMDs, this quirk emulates the VMLOCK and HPA shadow registers for all VMD device ids which don't natively assist with passthrough. The Linux VMD driver is updated to allow existing VMD devices to query VMLOCK for passthrough support.
Signed-off-by Jon Derrick <[email protected]>--- hw/vfio/pci-quirks.c | 103 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/pci.c | 7 +++ hw/vfio/pci.h | 2 + hw/vfio/trace-events | 3 ++ 4 files changed, 115 insertions(+)hw/vfio/pci-quirks.c +103
-0@@ -1709,3 +1709,106 @@ free_exit:return ret; } + +/* +* The VMD endpoint provides a real PCIe domain to the guest and the guest +* kernel performs enumeration of the VMD sub-device domain. Guest transactions +* to VMD sub-devices go through IOMMU translation from guest addresses to +* physical addresses. When MMIO goes to an endpoint after being translated to +* physical addresses, the bridge rejects the transaction because the window +* has been programmed with guest addresses. +* +* VMD can use the Host Physical Address in order to correctly program the +* bridge windows in its PCIe domain. VMD device 28C0 has HPA shadow registers +* located at offset 0x2000 in MEMBAR2 (BAR 4). The shadow registers are valid +* if bit 1 is set in the VMD VMLOCK config register 0x70. VMD devices without +* this native assistance can have these registers safely emulated as these +* registers are reserved. +*/ +typedef struct VFIOVMDQuirk { +VFIOPCIDevice *vdev; +uint64_t membar_phys[2 ]; +}VFIOVMDQuirk ; + +static uint64_t vfio_vmd_quirk_read (void *opaque,hwaddr addr,unsigned size) +{ +VFIOVMDQuirk *data = opaque; +uint64_t val =0 ; + +memcpy (&val, (void *)data->membar_phys + addr, size); +return val; +} + +static const MemoryRegionOps vfio_vmd_quirk = { + .read = vfio_vmd_quirk_read, + .endianness =DEVICE_LITTLE_ENDIAN , +}; + +#define VMD_VMLOCK 0x70 +#define VMD_SHADOW 0x2000 +#define VMD_MEMBAR2 4 + +static int vfio_vmd_emulate_shadow_registers (VFIOPCIDevice *vdev) +{ +VFIOQuirk *quirk; +VFIOVMDQuirk *data; +PCIDevice *pdev = &vdev->pdev; +int ret; + + data =g_malloc0 (sizeof (*data)); + ret =pread (vdev->vbasedev.fd, data->membar_phys,16 , + vdev->config_offset +PCI_BASE_ADDRESS_2 ); +if (ret !=16 ) { +error_report ("VMD %s cannot read MEMBARs (%d)" , + vdev->vbasedev.name, ret); +g_free (data); +return -EFAULT ; + } + + quirk =vfio_quirk_alloc (1 ); + quirk->data = data; + data->vdev = vdev; + +/* Emulate Shadow Registers */ +memory_region_init_io (quirk->mem,OBJECT (vdev), &vfio_vmd_quirk, data, +"vfio-vmd-quirk" ,sizeof (data->membar_phys)); +memory_region_add_subregion_overlap (vdev->bars[VMD_MEMBAR2 ].region.mem, +VMD_SHADOW , quirk->mem,1 ); +memory_region_set_readonly (quirk->mem, true); +memory_region_set_enabled (quirk->mem, true); + +QLIST_INSERT_HEAD (&vdev->bars[VMD_MEMBAR2 ].quirks, quirk, next); + +trace_vfio_pci_vmd_quirk_shadow_regs (vdev->vbasedev.name, + data->membar_phys[0 ], + data->membar_phys[1 ]); + +/* Advertise Shadow Register support */ +pci_byte_test_and_set_mask (pdev->config +VMD_VMLOCK ,0x2 ); +pci_set_byte (pdev->wmask +VMD_VMLOCK ,0 ); +pci_set_byte (vdev->emulated_config_bits +VMD_VMLOCK ,0x2 ); + +trace_vfio_pci_vmd_quirk_vmlock (vdev->vbasedev.name, +pci_get_byte (pdev->config +VMD_VMLOCK )); + +return 0 ; +} + +int vfio_pci_vmd_init (VFIOPCIDevice *vdev) +{ +int ret =0 ; + +switch (vdev->device_id) { +case 0x28C0 :/* Native passthrough support */ +break ; +/* Emulates Native passthrough support */ +case 0x201D : +case 0x467F : +case 0x4C3D : +case 0x9A0B : + ret =vfio_vmd_emulate_shadow_registers (vdev); +break ; + } + +return ret; +}hw/vfio/pci.c +7
-0@@ -3024,6 +3024,13 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) } } +if (vdev->vendor_id ==PCI_VENDOR_ID_INTEL ) { + ret =vfio_pci_vmd_init (vdev); +if (ret) { +error_report ("Failed to setup VMD" ); + } + } +vfio_register_err_notifier (vdev);vfio_register_req_notifier (vdev);vfio_setup_resetfn_quirk (vdev);hw/vfio/pci.h +2
-0@@ -217,6 +217,8 @@ int vfio_pci_igd_opregion_init(VFIOPCIDevice *vdev,int vfio_pci_nvidia_v100_ram_init (VFIOPCIDevice *vdev,Error **errp);int vfio_pci_nvlink2_init (VFIOPCIDevice *vdev,Error **errp); +int vfio_pci_vmd_init (VFIOPCIDevice *vdev); +void vfio_display_reset (VFIOPCIDevice *vdev);int vfio_display_probe (VFIOPCIDevice *vdev,Error **errp);void vfio_display_finalize (VFIOPCIDevice *vdev);hw/vfio/trace-events +3
-0@@ -90,6 +90,9 @@ vfio_pci_nvidia_gpu_setup_quirk(const char *name, uint64_t tgt, uint64_t size) " vfio_pci_nvlink2_setup_quirk_ssatgt(const char *name, uint64_t tgt, uint64_t size) "%s tgt=0x%"PRIx64" size=0x%"PRIx64 vfio_pci_nvlink2_setup_quirk_lnkspd(const char *name, uint32_t link_speed) "%s link_speed=0x%x" +vfio_pci_vmd_quirk_shadow_regs(const char *name, uint64_t mb1, uint64_t mb2) "%s membar1_phys=0x%"PRIx64" membar2_phys=0x%"PRIx64 +vfio_pci_vmd_quirk_vmlock(const char *name, uint8_t vmlock) "%s vmlock=0x%x" + # common.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64
Re: [PATCH for QEMU v2] hw/vfio: Add VMD Passthrough Quirk
Quoted 6 lines · Alex Williamson
Quoted 3 lines · Jonathan
Hi Alex,
[snip]
Quoted 11 lines
Thanks for the confirmation. The approach seems ok, but a subtle side-effect of MemoryRegion quirks is that we're going to trap the entire
PAGE_SIZErange overlapping the quirk out to QEMU on guest access. The two registers at0x2000might be reserved for this purpose, but is there anything else that could be performance sensitive anywhere in that page? If so, it might be less transparent, but more efficient to invent a new quirk making use of config space (ex. adding an emulated vendor capability somewhere known to be unused on the device). Thanks,
Quoted 1 line
Alex
Seems like that could be a problem if we're running with huge pages and overlapping msix tables.
Gerd also recommended adding a vendor specific capability. I don't see that there's any vendor-specific capabilities we can accidentally shadow, so I'll go ahead and do that.
Thanks,
Jon
Quoted 170 lines
Quoted 170 lines
Quoted 170 lines
Quoted 170 lines
In order to support existing VMDs, this quirk emulates the VMLOCK and HPA shadow registers for all VMD device ids which don't natively assist with passthrough. The Linux VMD driver is updated to allow existing VMD devices to query VMLOCK for passthrough support.
Signed-off-by Jon Derrick <[email protected]>--- hw/vfio/pci-quirks.c | 103 +++++++++++++++++++++++++++++++++++++++++++ hw/vfio/pci.c | 7 +++ hw/vfio/pci.h | 2 + hw/vfio/trace-events | 3 ++ 4 files changed, 115 insertions(+)hw/vfio/pci-quirks.c +103
-0@@ -1709,3 +1709,106 @@ free_exit:return ret; } + +/* +* The VMD endpoint provides a real PCIe domain to the guest and the guest +* kernel performs enumeration of the VMD sub-device domain. Guest transactions +* to VMD sub-devices go through IOMMU translation from guest addresses to +* physical addresses. When MMIO goes to an endpoint after being translated to +* physical addresses, the bridge rejects the transaction because the window +* has been programmed with guest addresses. +* +* VMD can use the Host Physical Address in order to correctly program the +* bridge windows in its PCIe domain. VMD device 28C0 has HPA shadow registers +* located at offset 0x2000 in MEMBAR2 (BAR 4). The shadow registers are valid +* if bit 1 is set in the VMD VMLOCK config register 0x70. VMD devices without +* this native assistance can have these registers safely emulated as these +* registers are reserved. +*/ +typedef struct VFIOVMDQuirk { +VFIOPCIDevice *vdev; +uint64_t membar_phys[2 ]; +}VFIOVMDQuirk ; + +static uint64_t vfio_vmd_quirk_read (void *opaque,hwaddr addr,unsigned size) +{ +VFIOVMDQuirk *data = opaque; +uint64_t val =0 ; + +memcpy (&val, (void *)data->membar_phys + addr, size); +return val; +} + +static const MemoryRegionOps vfio_vmd_quirk = { + .read = vfio_vmd_quirk_read, + .endianness =DEVICE_LITTLE_ENDIAN , +}; + +#define VMD_VMLOCK 0x70 +#define VMD_SHADOW 0x2000 +#define VMD_MEMBAR2 4 + +static int vfio_vmd_emulate_shadow_registers (VFIOPCIDevice *vdev) +{ +VFIOQuirk *quirk; +VFIOVMDQuirk *data; +PCIDevice *pdev = &vdev->pdev; +int ret; + + data =g_malloc0 (sizeof (*data)); + ret =pread (vdev->vbasedev.fd, data->membar_phys,16 , + vdev->config_offset +PCI_BASE_ADDRESS_2 ); +if (ret !=16 ) { +error_report ("VMD %s cannot read MEMBARs (%d)" , + vdev->vbasedev.name, ret); +g_free (data); +return -EFAULT ; + } + + quirk =vfio_quirk_alloc (1 ); + quirk->data = data; + data->vdev = vdev; + +/* Emulate Shadow Registers */ +memory_region_init_io (quirk->mem,OBJECT (vdev), &vfio_vmd_quirk, data, +"vfio-vmd-quirk" ,sizeof (data->membar_phys)); +memory_region_add_subregion_overlap (vdev->bars[VMD_MEMBAR2 ].region.mem, +VMD_SHADOW , quirk->mem,1 ); +memory_region_set_readonly (quirk->mem, true); +memory_region_set_enabled (quirk->mem, true); + +QLIST_INSERT_HEAD (&vdev->bars[VMD_MEMBAR2 ].quirks, quirk, next); + +trace_vfio_pci_vmd_quirk_shadow_regs (vdev->vbasedev.name, + data->membar_phys[0 ], + data->membar_phys[1 ]); + +/* Advertise Shadow Register support */ +pci_byte_test_and_set_mask (pdev->config +VMD_VMLOCK ,0x2 ); +pci_set_byte (pdev->wmask +VMD_VMLOCK ,0 ); +pci_set_byte (vdev->emulated_config_bits +VMD_VMLOCK ,0x2 ); + +trace_vfio_pci_vmd_quirk_vmlock (vdev->vbasedev.name, +pci_get_byte (pdev->config +VMD_VMLOCK )); + +return 0 ; +} + +int vfio_pci_vmd_init (VFIOPCIDevice *vdev) +{ +int ret =0 ; + +switch (vdev->device_id) { +case 0x28C0 :/* Native passthrough support */ +break ; +/* Emulates Native passthrough support */ +case 0x201D : +case 0x467F : +case 0x4C3D : +case 0x9A0B : + ret =vfio_vmd_emulate_shadow_registers (vdev); +break ; + } + +return ret; +}hw/vfio/pci.c +7
-0@@ -3024,6 +3024,13 @@ static void vfio_realize(PCIDevice *pdev, Error **errp) } } +if (vdev->vendor_id ==PCI_VENDOR_ID_INTEL ) { + ret =vfio_pci_vmd_init (vdev); +if (ret) { +error_report ("Failed to setup VMD" ); + } + } +vfio_register_err_notifier (vdev);vfio_register_req_notifier (vdev);vfio_setup_resetfn_quirk (vdev);hw/vfio/pci.h +2
-0@@ -217,6 +217,8 @@ int vfio_pci_igd_opregion_init(VFIOPCIDevice *vdev,int vfio_pci_nvidia_v100_ram_init (VFIOPCIDevice *vdev,Error **errp);int vfio_pci_nvlink2_init (VFIOPCIDevice *vdev,Error **errp); +int vfio_pci_vmd_init (VFIOPCIDevice *vdev); +void vfio_display_reset (VFIOPCIDevice *vdev);int vfio_display_probe (VFIOPCIDevice *vdev,Error **errp);void vfio_display_finalize (VFIOPCIDevice *vdev);hw/vfio/trace-events +3
-0@@ -90,6 +90,9 @@ vfio_pci_nvidia_gpu_setup_quirk(const char *name, uint64_t tgt, uint64_t size) " vfio_pci_nvlink2_setup_quirk_ssatgt(const char *name, uint64_t tgt, uint64_t size) "%s tgt=0x%"PRIx64" size=0x%"PRIx64 vfio_pci_nvlink2_setup_quirk_lnkspd(const char *name, uint32_t link_speed) "%s link_speed=0x%x" +vfio_pci_vmd_quirk_shadow_regs(const char *name, uint64_t mb1, uint64_t mb2) "%s membar1_phys=0x%"PRIx64" membar2_phys=0x%"PRIx64 +vfio_pci_vmd_quirk_vmlock(const char *name, uint8_t vmlock) "%s vmlock=0x%x" + # common.c vfio_region_write(const char *name, int index, uint64_t addr, uint64_t data, unsigned size) " (%s:region%d+0x%"PRIx64", 0x%"PRIx64 ", %d)" vfio_region_read(char *name, int index, uint64_t addr, unsigned size, uint64_t data) " (%s:region%d+0x%"PRIx64", %d) = 0x%"PRIx64
Re: [PATCH for QEMU v2] hw/vfio: Add VMD Passthrough Quirk
Quoted 20 lines · Jonathan
Quoted 6 lines · Alex Williamson
Quoted 3 lines · Jonathan
Hi Alex,
[snip]
Quoted 11 lines
Thanks for the confirmation. The approach seems ok, but a subtle side-effect of MemoryRegion quirks is that we're going to trap the entire
PAGE_SIZErange overlapping the quirk out to QEMU on guest access. The two registers at0x2000might be reserved for this purpose, but is there anything else that could be performance sensitive anywhere in that page? If so, it might be less transparent, but more efficient to invent a new quirk making use of config space (ex. adding an emulated vendor capability somewhere known to be unused on the device). Thanks,
Quoted 4 lines
Quoted 1 line
Alex
Seems like that could be a problem if we're running with huge pages and overlapping msix tables.
FWIW, MSI-X tables are already getting trapped into QEMU for emulation. We have a mechanism via x-msix-relocation=<OffAutoPCIBAR> in QEMU to deal with that when it becomes a problem by emulating the MSI-X structures in MMIO space that doesn't overlap with the actual device (ie. virtually resizing or adding BARs). The issue is what else can be in that 4K page or will this device be supported on archs where the system page size is >4K more so than huge pages (as in hugetlbfs or transparent huge pages). Thanks,
Alex
PATCH v2 1/2 PCI: vmd: Filter resource type bits from shadow register
In reply to Jon DerrickVersions of VMD with the Host Physical Address shadow register use this register to calculate the bus address offset needed to do guest passthrough of the domain. This register shadows the Host Physical Address registers including the resource type bits. After calculating the offset, the extra resource type bits lead to the VMD resources being over-provisioned at the front and under-provisioned at the back.
Example:
pci 10000:80:02.0: reg 0x10: [mem 0xf801fffc-0xf803fffb 64bit]
Expected:
pci 10000:80:02.0: reg 0x10: [mem 0xf8020000-0xf803ffff 64bit]
If other devices are mapped in the over-provisioned front, it could lead to resource conflict issues with VMD or those devices.
--- drivers/pci/controller/vmd.c | 6 ++++-- 1 file changed, 4 insertions(+), 2 deletions(-)
drivers/pci/controller/vmd.c +4 -2
@@ -445,9 +445,11 @@ static int vmd_enable_domain(struct vmd_dev *vmd, unsigned long features)if (!membar2)return -ENOMEM ; offset[0 ] = vmd->dev->resource[VMD_MEMBAR1 ].start - -readq (membar2 +MB2_SHADOW_OFFSET ); + (readq (membar2 +MB2_SHADOW_OFFSET ) & +PCI_BASE_ADDRESS_MEM_MASK ); offset[1 ] = vmd->dev->resource[VMD_MEMBAR2 ].start - -readq (membar2 +MB2_SHADOW_OFFSET +8 ); + (readq (membar2 +MB2_SHADOW_OFFSET +8 ) & +PCI_BASE_ADDRESS_MEM_MASK );pci_iounmap (vmd->dev, membar2); } }
Signature · 1 line
2.18.1
PATCH v2 2/2 PCI: vmd: Use Shadow MEMBAR registers for QEMU/KVM guests
In reply to Jon DerrickVMD device 28C0 natively assists guest passthrough of the VMD endpoint through the use of shadow registers that provide Host Physical Addresses to correctly assign bridge windows. These shadow registers are only available if VMD config space register 0x70, bit 1 is set.
For existing VMD which don't natively support the shadow register, VMD config space register 0x70 is reserved and will return 0. Future VMD will have these registers natively in hardware, but existing VMD can still use this feature by emulating the config space register and shadow registers.
QEMU has been modified to emulate this config space register and the shadow membar registers for VMDs which don't natively support this feature. This patch updates the supported device list to allow this feature to be used on these VMDs.
--- drivers/pci/controller/vmd.c | 15 ++++++++++----- 1 file changed, 10 insertions(+), 5 deletions(-)
drivers/pci/controller/vmd.c +10 -5
@@ -600,6 +600,7 @@ static irqreturn_t vmd_irq(int irq, void *data)static int vmd_probe (struct pci_dev *dev,const struct pci_device_id *id) {struct vmd_dev *vmd; +unsigned long features = id->driver_data;int i, err;if (resource_size (&dev->resource[VMD_CFGBAR ]) < (1 <<20 ))
@@ -652,7 +653,7 @@ static int vmd_probe(struct pci_dev *dev, const struct pci_device_id *id)spin_lock_init (&vmd->cfg_lock);pci_set_drvdata (dev, vmd); - err =vmd_enable_domain (vmd, (unsigned long ) id->driver_data); + err =vmd_enable_domain (vmd, features);if (err)return err;
@@ -716,16 +717,20 @@ static int vmd_resume(struct device *dev) static SIMPLE_DEV_PM_OPS(vmd_dev_pm_ops, vmd_suspend, vmd_resume);static const struct pci_device_id vmd_ids[] = { - {PCI_DEVICE (PCI_VENDOR_ID_INTEL ,PCI_DEVICE_ID_INTEL_VMD_201D ),}, + {PCI_DEVICE (PCI_VENDOR_ID_INTEL ,PCI_DEVICE_ID_INTEL_VMD_201D ), + .driver_data =VMD_FEAT_HAS_MEMBAR_SHADOW ,}, {PCI_DEVICE (PCI_VENDOR_ID_INTEL ,PCI_DEVICE_ID_INTEL_VMD_28C0 ), .driver_data =VMD_FEAT_HAS_MEMBAR_SHADOW |VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, {PCI_DEVICE (PCI_VENDOR_ID_INTEL ,0x467f ), - .driver_data =VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, + .driver_data =VMD_FEAT_HAS_MEMBAR_SHADOW | +VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, {PCI_DEVICE (PCI_VENDOR_ID_INTEL ,0x4c3d ), - .driver_data =VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, + .driver_data =VMD_FEAT_HAS_MEMBAR_SHADOW | +VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, {PCI_DEVICE (PCI_VENDOR_ID_INTEL ,PCI_DEVICE_ID_INTEL_VMD_9A0B ), - .driver_data =VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, + .driver_data =VMD_FEAT_HAS_MEMBAR_SHADOW | +VMD_FEAT_HAS_BUS_RESTRICTIONS ,}, {0 ,} };MODULE_DEVICE_TABLE (pci, vmd_ids);
Signature · 1 line
2.18.1