'struct nept_desc' is used to associate guest EPTP with a shadow EPTP.
It's created in the first reference and be freed while no reference.
The life cycle seems like,
While guest VMCS VMX_EPT_POINTER_FULL is changed, the 'struct nept_desc'
of the new guest EPTP is referenced; the 'struct nept_desc' of the old
guest EPTP is dereferenced.
While guest VMCS be cleared(by VMCLEAR in L1 VM), the 'struct nept_desc'
of the old guest EPTP is dereferenced.
While a new guest VMCS be loaded(by VMPTRLD in L1 VM), the 'struct
nept_desc' of the new guest EPTP is referenced. The 'struct nept_desc'
of the old guest EPTP is dereferenced.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
To shadow guest EPT, the hypervisor needs construct a shadow EPT for each
guest EPT. The key to associate a shadow EPT and a guest EPT is the EPTP
(EPT pointer). This patch provides following structure to do the association.
struct nept_desc {
/*
* A shadow EPTP.
* The format is same with 'EPT pointer' in VMCS.
* Its PML4 address field is a HVA of the hypervisor.
*/
uint64_t shadow_eptp;
/*
* An guest EPTP configured by L1 VM.
* The format is same with 'EPT pointer' in VMCS.
* Its PML4 address field is a GPA of the L1 VM.
*/
uint64_t guest_eptp;
uint32_t ref_count;
};
Due to lack of dynamic memory allocation of the hypervisor, a array
nept_bucket of type 'struct nept_desc' is introduced to store those
association information. A guest EPT might be shared between different
L2 vCPUs, so this patch provides several functions to handle the
reference of the structure.
Interface get_shadow_eptp() also is introduced. To find the shadow EPTP
of a specified guest EPTP.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Shadow EPT uses lots of pages to construct the shadow page table. To
utilize the memory more efficient, a page poll sept_page_pool is
introduced.
For simplicity, total platform RAM size is considered to calculate the
memory needed for shadow page tables. This is not an accurate upper
bound. This can satisfy typical use-cases where there is not a lot
of overcommitment and sharing of memory between L2 VMs.
Memory of the pool is marked as reserved from E820 table in early stage.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Nested VM exits happen when vCPU is in guest mode (VMCS02 is current).
Initially we reflect all nested VM exits to L1 hypervisor. To prepare
the environment to run L1 guest:
- restore some VMCS fields to the value as what L1 hypervisor programmed.
- VMCLEAR VMCS02, VMPTRLD VMCS01 and enable VMCS shadowing.
- load the non-shadowing host states from VMCS12 to VMCS01 guest states.
- VMRESUME to L1 guest with this modified VMCS01.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Signed-off-by: Alexander Merritt <alex.merritt@intel.com>
Since L2 guest vCPU mode and VPID are managed by L1 hypervisor, so we
can skip these handling in run_vcpu().
And be careful that we can't cache L2 registers in struct acrn_vcpu.
Tracked-On: #5923
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
invvpid and invept instructions cause VM exits unconditionally.
For initial support, we pass all the instruction operands as is
to the pCPU.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Implement the VMLAUNCH and VMRESUME instructions, allowing a L1
hypervisor to run nested guests.
- merge VMCS control fields and VMCS guest fields to VMCS02
- clear shadow VMCS indicator on VMCS02 and load VMCS02 as current
- set VMCS12 launch state to "launched" in VMLAUNCH handler
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Signed-off-by: Alex Merritt <alex.merritt@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Make up qemu.xml to compromise the static allocators which use the
new xpath based on board inspector.
Tracked-On: #6102
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
For illegal characters, replace original characters with escaped characters in board.xml.
Tracked-On: #6113
Signed-off-by: Kunhui Li <kunhuix.li@intel.com>
The latest version of RTCT specification is version 2.
This patch is to add RTCT v2 support for virtual RTCT
of guest.
Tracked-On: #6020
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
Emulation of guest lapic ID has been enhanced to
indicate the topology of vCPU hierarchy.
This patch refine logic to build virtual RCTC_v1 table
of guest to adapt above lapic ID changes.
Tracked-On: #6020
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
Signature of RTCT ACPI table maybe "PTCT"(v1) or "RTCT"(v2).
and the MAGIC number in CRL header is also changed from "PTCM"
to "RTCM".
This patch refine the code to detect RTCT table for both
v1 and v2.
Tracked-On: #6020
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Modify the initial value of PT_SLOT variable and
update the get slot logic that all device call the virtual_dev_slot function to get slot number directly.
Copy the launch_uos_id1.sh to launch_win.sh.
Tracked-On: #6072
Signed-off-by: Kunhui Li <kunhuix.li@intel.com>
Update the severity from "warning" to "error" for hybrid cores check.
Tracked-On: #5918
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Signed-off-by: Kunhui Li <kunhuix.li@intel.com>
We missed a section during our last update to the QEMU tutorial
that references which version (tag) of ACRN to use.
Tracked-On: #5928
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
hv is responsible to ensure PTM is always enabled on hw root port if
that root port is PTM root-capable. If PTM root is not enabled already in physical
root port before guest launch, guest OS can only enable it in root port's virtual
config space and PTM may not function as desired so we would rather not
to allow user to enable PTM on pass-thru device in this case.
Also revisit a few comments to add assumption that acrn only support a
simple PTM hierarch emulation i.e., EP (ptm requestor) is directly connected
to root port (ptm root), or ptm requestor is rcie).
V4:
- Change behavior from V3: When users tries to enable PTM while PTM root is not enabled on
physical root port, allow user to launch VM and pass thru the device
with no PTM support.
V3:
- When users tries to enable PTM while PTM root is not enabled on
physical root port, allow user to launch VM and enable PTM in virtual
root port. This is not going to enable PTM on physical root port.
Tracked-On: #5915
Signed-off-by: Rong Liu <rong.l.liu@intel.com>
Acked-by: Jason Chen <jason.cj.chen@intel.com>
For post launch VM, ACRN supports PTM under these conditions:
1. HW implements a simple PTM hierarchy: PTM requestor device (ep) is
directly connected to PTM root capable root port. Or
2. ptm requestor itself is root complex integrated ep.
Currently acrn doesn't support emulation of other type of PTM hiearchy, such
as if there is an intermediate PTM node (for example, switch) inbetween
PTM requestor and PTM root.
To avoid VM touching physical hardware, acrn hv ensures PTM is always enabled
in the hardware.
During hv's pci init, if root port is ptm capable,
hv will enable PTM on that root port. In addition,
log error (and don't enable PTM) if ptm root
capability is on intermediate node other than root port.
V2:
- Modify commit messages to clarify the limitation
of current PTM implementation.
- Fix code that may fail FUSA
- Remove pci_ptm_info() and put info log inside pci_enable_ptm_root().
Tracked-On: #5915
Signed-off-by: Rong Liu <rong.l.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
In physical destination mode, the destination processor is specified by its
local APIC ID. When a CPU switch xAPIC Mode to x2APIC Mode or vice versa,
the local APIC ID is not changed. So a vcpu in x2APIC Mode could use physical
Destination Mode to send an IPI to another vcpu in xAPIC Mode by writing ICR.
This patch adds support for a vCPU A could write ICR to send IPI to another
vCPU B which is in different APIC mode.
Tracked-On: #5923
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Retrieve physical APIC IDs from board xml file and use them to fill in the ACPI MADT table
for pre-Launched VMs.
Note that the config-tool will throw an error if the processors/die/core/thread tags are absent.
User needs to run board_inspector.py to regenerate the board xml file when this commit is merged,
if the processors/die/core/thread tags are missing in the board xml file.
Tracked-On: #6020
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
Two utility functions are copied and adapted from hyerpervisor:
ffs64
bitmap_clear_nolock
Two public functions are provided for future use (such as for RTCTv2)
pcpuid_from_vcpuid
lapicid_from_pcpuid
Tracked-On: #6020
Reviewed-by: Wang, Yu1 <yu1.wang@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
This allows users to retrieve and use the requested platform_info information from hypervisor
Tracked-On: #6020
Reviewed-by: Wang, Yu1 <yu1.wang@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
So that DM can retrieve physical APIC IDs and use them to fill in the ACPI MADT table for
post-launched VMs.
Note:
1. DM needs to use the same logic as hypervisor to calculate vLAPIC IDs based on physical APIC IDs
and CPU affinity setting
2. Using reserved0[] in struct hc_platform_info to pass physical APIC IDs means we can only support at
most 116 cores. And it assumes LAPIC ID is 8bits (X2APIC mode supports 32 bits).
Cat IDs shift will be used by DM RTCT V2
Tracked-On: #6020
Reviewed-by: Wang, Yu1 <yu1.wang@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
Using physical APIC IDs as vLAPIC IDs for pre-Launched and post-launched VMs
is not sufficient to replicate the host CPU and cache topologies in guest VMs,
we also need to passthrough host CPUID leaf.0BH to guest VMs, otherwise,
guest VMs may see weird CPU topology.
Note that in current code, ACRN has already passthroughed host cache CPUID
leaf 04H to guest VMs
Tracked-On: #6020
Reviewed-by: Wang, Yu1 <yu1.wang@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
In current code, ACRN uses physical APIC IDs as vLAPIC IDs for SOS,
and vCPU ids (contiguous) as vLAPIC IDs for pre-Launched and post-Launched VMs.
Using vCPU ids as vLAPIC IDs for pre-Launched and post-Launched VMs
would result in wrong CPU and cache topologies showing in the guest VMs,
and could adversely affect performance if the guest VM chooses to detect
CPU and cache topologies and optimize its behavior accordingly.
Uses physical APIC IDs as vLAPIC IDs (and related CPU/cache topology enumeration
CPUIDs passthrough) will replicate the host CPU and cache topologies in pre-Launched
and post-Launched VMs.
Tracked-On: #6020
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
Remove the direct calls to exec_vmptrld() or exec_vmclear(), and replace
with the wrapper APIs load_va_vmcs() and clear_va_vmcs().
Tracked-On: #5923
Signed-off-by: Zide Chen <zide.chen@intel.com>
commit 873ed75 ("misc: sanity check VM config for nested virtualization")
requires that the guest_flag tag can't be empty, or it will fail to build.
This patch changes adl instances of "<guest_flag></guest_flag>" to
"<guest_flag>0</guest_flag>".
Tracked-On: #5923
Signed-off-by: Jiang, Yanting <yanting.jiang@intel.com>
Configure PTM in post-launched VM using <PTM> element. If the //vm/PTM
sets to 'y', pci_dev.c.xsl appends the virtual root port to
corresponding struct acrn_vm_pci_dev_config of that VM. Currently it
supports only post-launched VMs.
Configure enable_ptm for dm argument. If a uos/enable_ptm with uos id
= 'vm_id 'sets to 'y' and the vm/PTM with the same vm_id sets to 'y',
append an "enable_ptm" flag to the end of passthrough ethernet devices.
Currently there is only ethernet card can support the "enable_ptm"flag.
For the schema validation, the <PTM> can only be ['y', 'n'].
For the launched script validation, the <enable_ptm> can only be ['y',
'n']. If the <enable_ptm> sets to 'y' but the corresponding <PTM> sets
to 'n', the launch script will fail to generate.
Tracked-On: #6054
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Add a number of steps and details that were not called
out in the "Getting Started Guide for ACRN Hybrid Mode". Those
are not obvious to the first-time or novice user so the user
guide was hard to follow and confusing. At a high-level:
* How to build Zephyr
* How to install ACRN
* How to install the ACRN kernel
The hybrid scenario overview diagram has been updated too.
Tracked-On: #5992
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
Add an xslt file "board_info.h.xsl". This file is used to
generate board_info.h which is used by hypervisor.
Tracked-On: #6024
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add an xslt file "pci_dev.c.xsl". This file is used to
generate pci_dev.c which is used by hypervisor.
Tracked-On: #6024
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
acrn:get-vbdf: get the virtual bdf from allocation.xml based on vmid and device name
acrn:get-pbdf: get physical bdf from <pci_dev>
acrn:ptdev-name-suffix: fix the name to look up allocation.xml
acrn:get-hidden-device-num: get the number of hidden devices based on
board name
acrn:is-vmsix-supported-device: check if a device is a vmsix supported
device based on the vendor and identifier
Tracked-On: #6024
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Assign bdf to pci emulated and passthrough devices.
For pre-launched VM, assigns unique bdf to passthrough devices, inter-vm
shared memory, pci vuart(console and communication vuarts).
For SOS vm, assigns unique bdf to inter-vm shared memory and pci
vuart(console and communication vuarts).
The bdf follows the rules below:
- the bdf 00:00.0 is reserved for pci hostbridge
- the assigned bdf range: bus is 0x00, dev is in range [0x1, 0x20)
and the fuc is 0x00
- the bdf must be unique, which means any vm's emulated devices cannot
share the same bdf with existing devices
- some devices's bdf is hardcoded, modify its bdf would leads the
device cannot be dicoverd by os. A HARDCODED_BDF_LIST in bdf.py documents
them
- the passthrough devices' bdf can be reused in SOS vm
Tracked-On: #6024
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add methods allocates the mmio bar base to console vuart,
communication vuarts, inter-vm shared memory and passthrough pci
devices.
For SOS:
- get low mem by parsing board xml.
- get high mem by parsing board xml, if the high mem is not enabled,
the high mem start address would be ~0UL and the end address is 0UL
- get the occupied mmio windows by parsing board.xml
- for each console vuart, communication vuart and inter-vm shared memory
devices, assign unused mmio windows to them
- all the assigned mmio windows must be unique and should not overlay
with any devices' mmio window
- the passthrough devices mmio windows can be reused in SOS vm
- each allocated mmio start address must be 4k alignment if the length
of bar is smaller than 4k
- each allocated mmio start address must be aligned with the bar length
if its length is greater than 4k
- the 32bits bar will fall in low mem range only
- 64bits bar will look for free mmio in low mem rage first, if the high
mem is enabled, the 64bits bar will look for free mmio in high mem
range if there is not enough space in low mem range
- allocator raises an error if there is not enough mmio space
For pre-launched VM:
- the high mem range is [256G, 512G)
- the low mem range is [2G, 3.5G)
- there is no used mmio window initially
- for each console vuart, communication vuart, inter-vm shared memory
devices and passthrough devices, assign unused mmio windows to them
- all the assigned mmio windows must be unique and should not overlay
with any devices' mmio window
- the 32bits bar will fall in low mem range only
- 64bits bar will look for free mmio in low mem rage first and then
look for free mmio in high mem range if there is not enough space in
low mem range
- each allocated mmio start address must be 4k alignment if the length
of bar is smaller than 4k
- each allocated mmio start address must be aligned with the bar length
if its lenght is greater than 4k
- allocator raises an error if there is not enough mmio space
Tracked-On: #6024
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add a common method "get_shmem_regions":
This method get <IVSHMEM_REGION> and extracts the region size, region
position in xml and and vm ids which share this regions. Returns a
dictionary:
{'vm_id':{'region_name':{'id': region position,'size': region size,}}}
Add vm type checking methods:
is_pre_launched_vm, is_post_launched_vm and is_sos_vm.
Tracked-On: #6024
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Only free rb_entry when we remove this entry from the rb tree, otherwise, a
page fault would trigger when next rb itreation would access the freed rb_entry.
Tracked-On: #6056
Signed-off-by: Li Fei1 <fei1.li@intel.com>
This patch implements the VMREAD and VMWRITE instructions.
When L1 guest is running with an active VMCS12, the “VMCS shadowing”
VM-execution control is always set to 1 in VMCS01. Thus the possible
behavior of VMREAD or VMWRITE from L1 could be:
- It causes a VM exit to L0 if the bit corresponds to the target VMCS
field in the VMREAD bitmap or VMWRITE bitmap is set to 1.
- It accesses the VMCS referenced by VMCS01 link pointer (VMCS02 in
our case) if the above mentioned bit is set to 0.
This patch handles the VMREAD and VMWRITE VM exits in this way:
- on VMWRITE, it writes the desired VMCS value to the respective field
in the cached VMCS12. For VMCS fields that need to be synced to VMCS02,
sets the corresponding dirty flag.
- on VMREAD, it reads the desired VMCS value from the cached VMCS12.
Tracked-On: #5923
Signed-off-by: Alex Merritt <alex.merritt@intel.com>
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@Intel.com>
This patch is to emulate VMCLEAR instruction.
L1 hypervisor issues VMCLEAR on a VMCS12 whose state could be any of
these: active and current, active but not current, not yet VMPTRLDed.
To emulate the VMCLEAR instruction, ACRN sets the VMCS12 launch state to
"clear", and if L0 already cached this VMCS12, need to sync it back to
guest memory:
- sync shadow fields from shadow VMCS VMCS to cache VMCS12
- copy cache VMCS12 to L1 guest memory
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Enable VMCS shadowing for most of the VMCS fields, so that execution of
the VMREAD or VMWRITE on these shadow VMCS fields from L1 hypervisor
won't cause VM exits, but read from or write to the shadow VMCS.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Alexander Merritt <alex.merritt@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Software layout of VMCS12 data is a contract between L1 guest and L0
hypervisor to run a L2 guest.
ACRN hypervisor caches the VMCS12 which is passed down from L1 hypervisor
by the VMPTRLD instructin. At the time of VMCLEAR, ACRN syncs the cached
VMCS12 back to L1 guest memory.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@Intel.com>
This patch emulates the VMPTRLD instruction. L0 hypervisor (ACRN) caches
the VMCS12 that is passed down from the VMPTRLD instruction, and merges it
with VMCS01 to create VMCS02 to run the nested VM.
- Currently ACRN can't cache multiple VMCS12 on one vCPU, so it needs to
flushes active but not current VMCS12s to L1 guest.
- ACRN creates VMCS02 to run nested VM based on VMCS12:
1) copy VMCS12 from guest memory to the per vCPU cache VMCS12
2) initialize VMCS02 revision ID and host-state area
3) load shadow fields from cache VMCS12 to VMCS02
4) enable VMCS shadowing before L1 Vm entry
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
This patch implements the VMXOFF instruction. By issuing VMXOFF,
L1 guest Leaves VMX Operation.
- cleanup VCPU nested virtualization context states in VMXOFF handler.
- implement check_vmx_permission() to check permission for VMX operation
for VMXOFF and other VMX instructions.
Tracked-On: #5923
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@Intel.com>
