Virtual interrupts injection and memory mapping operations can impact
target VM. By design, these type of operations from lower severity VM
to higher severity VM should be blocked by the hypervisor.
While the hypercalls are the interface between SOS VM and the
hypervisor, severity checks can be implemented at the beginning of
hypercalls needed.
Added severity checks in below hypercalls:
* hcall_set_vm_memory_regions()
* hcall_notify_ioreq_finish()
* hcall_set_irqline()
* hcall_inject_msi()
* hcall_write_protect_page()
Tracked-On: #5208
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
This patch is used to create and destroy an ivshmem device which
is emulated in hypervisor.
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Wang, Yu1 <yu1.wang@intel.com>
add create_ivshmem_from_dm and destroy_ivshmem_from_dm for
ivshmem device creation and destruction in dm-land
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Wang, Yu1 <yu1.wang@intel.com>
Add IC_CREATE_DEVICE and IC_DESTROY_DEVICE ioctls to create and
destroy an emulated device in hypervisor
v3: change IC_CREATE_DEVICE and IC_DESTROY_DEVICE to IC_CREATE_HV_VDEV
and IC_DESTROY_HV_VDEV
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Wang, Yu1 <yu1.wang@intel.com>
if device configuration vbdf is unassigned, then the corresponding
vdev will not be initialized, instead, the vdev will be initialized
by device model through hypercall.
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
For ivshmem vdev creation, the vdev vBDF, vBARs, shared memory region
name and size are set by device model. The shared memory name and size
must be same as the corresponding device configuration which is configured
by offline tool.
v3: add a comment to the vbar_base member of the acrn_vm_pci_dev_config
structure that vbar_base is power-on default value
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Add HC_CREATE_VDEV and HC_DESTROY_VDEV two hypercalls that are used to
create and destroy an emulated device(PCI device or legacy device) in hypervisor
v3: 1) change HC_CREATE_DEVICE and HC_DESTROY_DEVICE to HC_CREATE_VDEV
and HC_DESTROY_VDEV
2) refine code style
v4: 1) remove unnecessary parameter
2) add VM state check for HC_CREATE_VDEV and HC_DESTROY hypercalls
Tracked-On: #4853
Reviewed-by: Wang, Yu1 <yu1.wang@intel.com>
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
1.Modify clos_mask and mba_delay as a member of the union type.
2.Move HV_SUPPORTED_MAX_CLOS ,MAX_CACHE_CLOS_NUM_ENTRIES and
MAX_MBA_CLOS_NUM_ENTRIES to misc_cfg.h file.
Tracked-On: #5229
Signed-off-by: Wei Liu <weix.w.liu@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
1.Add macro MAX_CACHE_CLOS_NUM_ENTRIES for CAT, and MAX_MBA_CLOS_NUM_ENTRIES for MBA.
MAX_MBA_CLOS_NUM_ENTRIES:
Max number of Cache Mask entries corresponding to each CLOS.
This can vary if CDP is enabled vs disabled, as each CLOS entry will have corresponding
cache mask values for Data and Code when CDP is enabled.
MAX_CACHE_CLOS_NUM_ENTRIES:
Max number of MBA delay entries corresponding to each CLOS.
2.Move VMx_VCPU_CLOS macro to misc_cfg.h head file.
Tracked-On: #5229
Signed-off-by: Wei Liu <weix.w.liu@intel.com>
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
HV_SUPPORTED_MAX_CLOS:
This value represents the maximum CLOS that is allowed by ACRN hypervisor.
This value is set to be least common Max CLOS (CPUID.(EAX=0x10,ECX=ResID):EDX[15:0])
among all supported RDT resources in the platform. In other words, it is
min(maximum CLOS of L2, L3 and MBA). This is done in order to have consistent
CLOS allocations between all the RDT resources.
Tracked-On: #5229
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
New board, EHL CRB, does not have legacy port IO UART. Even the PCI UART
are not work due to BIOS's bug workaround(the BARs on LPSS PCI are reset
after BIOS hand over control to OS). For ACRN console usage, expose the
debug UART via ACPI PnP device (access by MMIO) and add support in
hypervisor debug code.
Another special thing is that register width of UART of EHL CRB is
1byte. Introduce reg_width for each struct console_uart.
Tracked-On: #4937
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
add IVSHMEM config in hybrid_rt scenario on tgl-rvp board.
Tracked-On: #4853
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Acked-by: Victor Sun <victor.sun@intel.com>
We use sos kernel cmdline maxcpus to limit the pCPU number of SOS
for hybrid or hybrid_rt scenarios by vcpu numbers calculation.
v2: add SOS CPU affinity calculation by total pCPU plus pCPUs
occupied by pre-launched VMs when no pcpuid configured
in SOS.
Tracked-On: #5216
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Reviewed-by: Victor Sun <victor.sun@intel.com>
Also clear Linux is no longer supported either as SOS or post-launched VM kernel.
- When it mentions clear Linux, mostly replaced by Ubuntu.
- remove all contents re/lated to "UEFI boot".
- remove the term de-privilege mode, and direct mode as well.
Tracked-On: #5197
Signed-off-by: Zide Chen <zide.chen@intel.com>
-- use an array to fast locate the hypercall handler
to replace switch case.
-- uniform hypercall handler as below:
int32_t (*handler)(sos_vm, target_vm, param1, param2)
Tracked-On: #4958
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Reviewed-by: Eddie Dong <eddie.dong@intel.com>
Add cfl-k700-i7 board xml and its industry xml to support ACRN industry
scenario on cfl-k700-i7 board.
Tracked-On: #5212
Signed-off-by: Victor Sun <victor.sun@intel.com>
Enhance the help text that accompanies the CONFIG_SCENARIO symbol in Kconfig
Tracked-On: #5203
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
CPU sharing between pre-launch VMs and SOS, post-launch VMs were
forbidden.
Remove the limitation.
Tracked-On: #5153
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Acked-by: Terry Zou <terry.zou@intel.com>
Add description of the new Hybrid RT (hybrid_rt) scenario to the "What is ACRN"
documentation.
Tracked-On: #5203
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
This patch is to support the Inter-VM communication by IVSHMEM
in config tool.
Users can configure IVSHMEM_ENABLE to enable or disable Inter-VM
communication by IVSHMEM; users can configure the name, size,
communication VM IDs of the IVSHMEM devices in the VM settings of
scenario xmls, then config tool will generate the related IVSHMEM
configurations for Inter-VM communication.
The config tool will do sanity check including when saving the xmls:
the format of shared memory region configuration is
[name],[size],[VM ID]:[VM ID](:[VM ID]...);
the max size of the name is 32 bytes;
the names should not be duplicated;
the mininum value of shared memory region size is 2M;
the value of shared memory region is a power of 2;
the size of share memory region should not extended the size of
available ram.
Tracked-On: #4853
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Acked-by: Victor Sun <victor.sun@intel.com>
Shared memory regoins can be added or deleted or updated from
scenario settings in config app with sanity check.
v2: move IVSHMEM config to hv->FEATURES->IVSHMEM
Tracked-On: #4853
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Acked-by: Victor Sun <victor.sun@intel.com>
add IVSHMEM_ENABLED and IVSHMEM_REGION in scenario xmls to support
Inter-VM communications configuration for VMs.
v2: move IVSHMEM config into <FEATURES> section
Tracked-On: #4853
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Acked-by: Victor Sun <victor.sun@intel.com>
Enable TPM passthrough configuration for pre-launched VM feature, on
TGL, by adding 'tgl-rvp' to TPM_PASSTHRU_BOARD.
Tracked-On: #5205
Signed-off-by: Tao Yuhong <yuhong.tao@intel.com>
Get the max number with integer list to instead string 'number'.
Tracked-On: #5199
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Acked-by: Victor Sun <victor.sun@intel.com>
Add more explanations about the various ACRN modes of operation and how they are
used to construct more complex scenarios.
Tracked-On: #5196
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
2abbb99f6a ("hv: make thread status more accurate") introduced a
transition stage, marked as var be_blocking, between RUNNING->BLOCKED
of thread status. wake_thread() does not work in this transition stage
because it only checks thread->status.
Need to check thread->be_blocking as well in wake_thread(). When
wake_thread() happens in the transition stage, the previous sleep
operation rolled back.
Tracked-On: #5190
Fixes: 2abbb99f6a ("hv: make thread status more accurate")
Signed-off-by: Conghui Chen <conghui.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
The default memory is 16G on TGL; the value of PLATFORM_RAM_SIZE and
SOS_RAM_SIZE is a little small in default xml.
Tracked-On: #5184
Reviewed-by: Victor Sun <victor.sun@intel.com>
Signed-off-by: fuzhongl <fuzhong.liu@intel.com>
- add a rule for pointer arithmetic operation
- add a rule about ABI conformance
- update GCC reference
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
Replace pr_fatal with pr_info to reduce printing logs
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Ivshmem device defines four registers including Interrupt Mask, Interrupt
Status, IVPostion and Doorbell. The first two are useless and no emulation
is required. The latter two are used for interrupts and will be implemented
in the future.
This patch also introduces a new priv_data member for structure pci_vdev,
it can be used to find an ivshmem device through pci_vdev.
v2: refine code style
v3: 1) add @pre for ivshmem_mmio_handler function
2) refine code style
v4: 1) set ivshmem registers default value when vBAR mapping
2) change find_ivshmem_device to set_ivshmem_device
v5: 1) change set_ivshmem_device to find_and_set_ivshmem_device
2) add a ASSERT to check if the vdev->priv_data is set successfully
v6: change find_and_set_ivshmem_device to create_ivshmem_device
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Implement read_vdev_cfg/write_vdev_cfg operations for ivshmem deivce
v2: read_vdev_cfg/write_vdev_cfg always return zero, the ivshmem device
only emulated in HV.
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
This patch introduces vpci_update_one_vbar API to simplify
vBAR mapping/unmapping when vBAR writing.
v2: refine commit message
v4: refine commit message
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
ivshmem device supports two BARs, BAR 0 is used for inter-VM
notification mechanism, BAR 2 is used to provide shared memory
base address and size.
v4: check if the return value of get_shm_region function is NULL
v5: 1) change get_shm_region to find_shm_region
2) add print log when ivshmem device doesn't find memory region
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
add an IVSHMEM regoin and the related configuration parameters in
hybrid_rt scenario on whl-ipc-i5. The size of the shared memory is
2M, and it is used for the communication between VM0 and VM2.
v6: rename shm name; remove unnecessary MACROs.
v7: rename MACRO for shm name; add unassigned vbdf for post-launched
VMs.
Tracked-On: #4853
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Ivshmem device is used for shared memory based communication between
pre-launched/post-launched VMs.
this patch implements ivshmem device configuration space initialization
and ivshmem device operation methods.
v2: introduce init_one_pcibar interface to simplify BAR initialization
operation of HV emulated PCI device.
v3: 1) due to init_one_pcibar API is only used for pre-launched VM vdevs
it can't be applied to all vdevs, so remove it.
2) move ivshmem BARs initialization to subsequent patch, this patch
only introduce ivshmem configuration space initialization.
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
The ivshmem memory regions use the memory of the hypervisor and
they are continuous and page aligned.
this patch is used to initialize each memory region hpa.
v2: 1) if CONFIG_IVSHMEM_SHARED_MEMORY_ENABLED is not defined, the
entire code of ivshmem will not be compiled.
2) change ivshmem shared memory unit from byte to page to avoid
misconfiguration.
3) add ivshmem configuration and vm configuration references
v3: 1) change CONFIG_IVSHMEM_SHARED_MEMORY_ENABLED to CONFIG_IVSHMEM_ENABLED
2) remove the ivshmem configuration sample, offline tool provides default
ivshmem configuration.
3) refine code style.
v4: 1) make ivshmem_base 2M aligned.
Tracked-On: #4853
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Fix build issue while CDP_ENABLED=y for EHL-CRB-B.
Tracked-On: #5092
Signed-off-by: Wei Liu <weix.w.liu@intel.com>
Acked-by: Victor Sun <victor.sun@intel.com>
There's a corner case:
When want to get CPUID.01H:EDX value,
may have the following code snippet:
uint32_t unused,edx;
cpuid_subleaf(0x1U, 0x0U, &unused, &unused, &unused, &edx);
while in cpuid_subleaf:
*eax = leaf;
*ecx = subleaf;
eax and ecx point to the same location,
When deep into asm_cpuid, it's input value will be 0x0U and 0x0U.
but the expected input value is 0x1U and 0x0U.
This case will return CPUID.00H:EDX, which is the wrong answer.
Tracked-On: #4526
Signed-off-by: Junming Liu <junming.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
The default memory is 16G on TGL; the value of HV and sos
ramsize is a little small in default xml.
Tracked-On: # 5184
Signed-off-by: fuzhongl <fuzhong.liu@intel.com>
Shuo A Liu