Some OS like Zephyr need to run at 0x100000, so modify the ve820 table
accordingly;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Previously multiboot mods[0] is designed for kernel module for all
pre-launched VMs including SOS VM, and mods[0].mm_string is used
to store kernel cmdline. This design could not satisfy the requirement
of hybrid mode scenarios that each VM might use their own kernel image
also ramdisk image. To resolve this problem, we will use a tag in
mods mm_string field to specify the module type. If the tag could
be matched with os_config of VM configurations, the corresponding
module would be loaded;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Previously the bootargs of SOS_VM is stored in a text file and stitched
into multiboot mods[0].string whereas the bootargs of PRE_LAUNCHED_VM is
stored in vm_configurations.c. Given the mods[].string will be used to
store Kernel image signature under hybrid mode, move the bootargs of SOS_VM
to vm configurations also to make it consistent with PRE_LAUNCHED_VM;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Use a misc_cfg.h in each board configs folder so that VM configurations
could include board specific MACROs;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Different kernel has different load method, it should be configurable
in vm configurations;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
For a Linux guest, the kernel load address should be gotten from zeropage
even for a pre-launched VM;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Currently the algorithm of direct_boot_sw_loader() is Linux bzImage specific,
so separate the bzImage specific loader code from it to make the api more
generic for other OS;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
The guest OS of ACRN will not be limited to Linux, so refine the struct
of sw_linux to more generic sw_module_info. Currently bootargs and ramdisk
are only supported modules but we can include more modules in future;
Tracked-On: #3214
Signed-off-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Ignore SIGPIPE signal and handle the error directly when write()
function fails. this will help us to catch the write failure
rather than crashing the UOS.
PS: Observed that when acrnctl stop command was issued from
kata-runtime, UOS was crashing with SIGPIPE signal.
Tracked-On: #3190
Signed-off-by: Vijay Dhanraj <vijay.dhanraj@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
When virtio-console is used as console port with socket backend,
guest kernel tries to hook it up with hvc console and sets it up.
It doesn't check if a client is connected and can result in ENOTCONN
with virtio-console backend being reset. This will prevent client
connection at a later point. To avoid this, ignore ENOTCONN error.
PS: For Kata, the runtime first launches VM and then proxy which acts
as a client connects to this socket. If this error is not handled,
proxy will never be able to connect as the backend itself will be reset.
Tracked-On: #3189
Signed-off-by: Vijay Dhanraj <vijay.dhanraj@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
Add vuart_deinit to vm shutdown so that the vuart resource can be
reset, and when the Guest VM restart, it could have right state.
Tracked-On: #2987
Signed-off-by: Conghui Chen <conghui.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
When enable Transmitter Holding Register Empty interrupts, vuart should
trigger interrupt when the THR is empty. In the previous code, only
after the data is written to THR, the flag thre_int_pending is set to
true. The thre_int_pending should also be true after THRE is set.
Besides, add logic to make sure interrupt is in right state when vuart
is initiated.
Tracked-On: #2987
Signed-off-by: Conghui Chen <conghui.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Some physical bar sizes are not page aligned, in order to support them,
we map a bigger region which is page aligned.
Tracked-On: #3181
Signed-off-by: Tw <wei.tan@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
PCI spec says that the procedure of sizing a BAR is as follows:
1) disable the decode via command register
2) save the original value of BAR register
3) write all-1 to the BAR register
4) read BAR register back, calculate the size
5) restore the original value to BAR register
6) re-enable the decode via command register
Some driver does not disable the decode of BAR register via the
command register before sizing a BAR. This will lead to a overlay
of the BAR addresses when trying to register the intermediate
BAR address via register_bar. A stateful variable sizing is used
to keep track of such kind of BAR address changes and workaroud
this violation.
Currently this issue is only found when audio device is passed
through to Windows 10 guest. When it is fixed in the Windows
audio driver, this patch should be reverted.
v1 -> v2:
- change the commit message to add the procedure of BAR sizing from
PCI spec
Tracked-On: #2962
Signed-off-by: Jian Jun Chen <jian.jun.chen@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
This patch is to fix error logs in client sending to server via socket.
Since gdb inherits the socket descriptor of client which is used as its
stderr, in this case, the error logs from gdb will mess up server, so
closing the inheritance when the socket is created.
Tracked-On: #3200
Signed-off-by: xiaojin2 <xiaojing.liu@intel.com>
Reviewed-by: Liu, Xinwu <xinwu.liu@intel.com>
Acked-by: Chen, Gang <gang.c.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
In vxworks, virtio-console FE driver only initiate 2 virtqueues, but BE
creates 2+ virtqueues for it. So the rest of the virtqueues are not
initiated. vq->used->flags cannot be used directly without any
condition.
Tracked-On: #3203
Signed-off-by: Gao Junhao <junhao.gao@intel.com>
Reviewed-by: Yu Wang <yu1.wang@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
is_msix, part of ptirq_msi_info, is not used in the code.
Tracked-On: #3205
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Since ACPI 2.0 a new field has been added to the table, X_FirmwareControl of type GAS,
which is 64bit. So use it conditionally.
Tracked-On: #3198
Signed-off-by: Tw <wei.tan@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
For PCI passthrough device when guest OS updates the BAR address
the corresponding EPT mapping should be updated as well.
Tracked-On: #2962
Signed-off-by: Jian Jun Chen <jian.jun.chen@intel.com>
Signed-off-by: Liu Shuo A <shuo.a.liu@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Yin Fengwei <fengwei.yin@intel.com>
Currently when get_rsdp is called, the EFI depriv_boot env is not initialized.
In such case it will fallback to the legacy mechanism of ACPI table.
If the ACPI table based on legacy mechanism is not found, it will fail to get
the ACPI table and then the system will hang.
On the old platform it still can parse the ACPI table from legacy mechanism.
In fact when EFI RSDP exists, the EFI RSDP is preferred instead of legacy ACPI
RSDP.
In order to avoid multiple calling of depriv_init_boot, the init_boot_operations
is renamed and called after X2apic is enabled(early_init_lapic).
Tracked-On: #3184
Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Reviewed-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Currently the get_rsdp() will be called several times and on some
platform it needs to scan the signature of "RSP ptr" in BIOS EBDA region
and 0xE0000-0xFFFFF region. It is slow.
So after the RSDP is parsed, it will be assigned to acpi_rsdp.
Tracked-On: #3184
Signed-off-by: Zhao Yakui <yakui.zhao@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
According to Chap 23.8 RESTRICTIONS ON VMX OPERATION, Vol 3, SDM:
"Any attempt to set one of these bits to an unsupported value while in VMX
operation (including VMX root operation) using any of the CLTS, LMSW, or
MOV CR instructions causes a general-protection exception."
So we don't need to trap them out then inject the GP in hypervisor.
Tracked-On: #2561
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
FuSa requires setting CR4.SMAP/SMEP/PKE will invalidate the TLB. However,
setting CR4.SMAP will invalidate the TLB on native while not in non-root mode.
To make sure this, we will trap CR4.SMAP/SMEP/PKE setting to invalidate the TLB
in root mode.
Tracked-On: #2561
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Since the vapic_id is from VM, need to check for pre-condition before passing
vcpu_id to vcpu_from_vid. This is in the path of LAPIC passthrough ICR
access
Tracked-On: #3170
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Clean vmcfg from doc/developer-guides/hld/hld-devicemodel.rst
Tracked-On: #3192
Acked-by: Anthony Xu <anthony.xu@intel.com>
Signed-off-by: Tao Yuhong <yuhong.tao@intel.com>
Removed the entire vmcfg feature, these changes are made:
1.Deleted devicemodel/vmcfg/
2.Deleted devicemodel/include/vmcfg.h
3.Cleanup vmcfg in devicemodel/Makefile
Tracked-On: #3192
Acked-by: Anthony Xu <anthony.xu@intel.com>
Signed-off-by: Tao Yuhong <yuhong.tao@intel.com>
Vmcfg is useless and should be removed. Before removing vmcfg code,
cleanup vmcfg APIs in core/main.c
Tracked-On: #3192
Acked-by: Anthony Xu <anthony.xu@intel.com>
Signed-off-by: Tao Yuhong <yuhong.tao@intel.com>
Reserved 128MB memory range for EPC resource in E820 table, starting
from 0x80000000.
Need to align the base address b/t DM and HV.
For hypervisor, the base address will be specified in epc field in
vm_configurations.c
Tracked-On: #3179
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Present SGX related MSRs to guest if SGX is supported.
- MSR_IA32_SGXLEPUBKEYHASH0 ~ MSR_IA32_SGXLEPUBKEYHASH3:
SGX Launch Control is not supported, so these MSRs are read only.
- MSR_IA32_SGX_SVN_STATUS:
read only
- MSR_IA32_FEATURE_CONTROL:
If SGX is support in VM, opt-in SGX in this MSR.
- MSR_SGXOWNEREPOCH0 ~ MSR_SGXOWNEREPOCH1:
The two MSRs' scope is package level, not allow guest to change them.
Still leave them in unsupported_msrs array.
Tracked-On: #3179
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
If sgx is supported in guest, present SGX capabilities to guest.
There will be only one EPC section presented to guest, even if EPC
memory for a guest is from muiltiple physcial EPC sections.
Tracked-On: #3179
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Reviewed-by: Yan, Like <like.yan@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Build EPT entries for SGX EPC resource for VMs.
- SOS: EPC resrouce will be removed from EPT of SOS, don't support SGX virtualization for SOS.
- Non-SOS: build ept mapping for EPC resource for guest.
Guest base address and size is specified in vm configuration.
Tracked-On: #3179
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Add EPC information in vm configuration structure.
EPC information contains the EPC base and size allocated to a VM.
Tracked-On: #3179
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Get the platform EPC resource and partiton the EPC resource for VMs
according to VM configurations.
Don't support sgx capability in SOS VM.
init_sgx is called during platform bsp initialization.
If init_sgx() fails, consider it as configuration error, panic the system.
init_sgx() fails if one of the following happens when at least one VM requests
EPC resource if no enough EPC resource for all VMs.
No further check if sgx is not supported by platform or not opted-in in BIOS,
just disable SGX support for VMs.
Tracked-On: #3179
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
When we call reset_vm() function to reset vm, the vm state
should be reset to VM_CREATED as well.
Tracked-On: #3182
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Fix a logic error when set iommu page walk coherent.
Tracked-On: #3160
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
VM Name length is restricted to 32 characters. kata creates
a VM name with GUID added as a part of VM name making it around
80 characters. So increasing this size to 128.
v1->v2:
It turns out that MAX_VM_OS_NAME_LEN usage in DM and HV are for
different use cases. So removing the macro from acrn_common.h.
Definied macro MAX_VMNAME_LEN for DM purposes in dm.h. Retaining
original macron name MAX_VM_OS_NAME_LEN for HV purposes but defined
in vm_config.h.
Tracked-On: #3138
Signed-off-by: Vijay Dhanraj <vijay.dhanraj@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Currently, we use STR_LEN for all checking the size of all the
acrn-dm parameters. But some parameters like kernel boot args
can grow based on different needs. For example, when kata launches
guest VM using acrn, the kernel boot args increases by 256 bytes
(i.e 1024 +256).
Just increasing STR_LEN will unnecessarily increase allocations
for other acrn-dm parameters. So decoupling only boot_args
length and increasing it to 2048.
PS: If other parameters like ramdisk path, kernel path,
elf_path etc. don't need 1024 bytes, we can reduce STR_LEN
to 256 or 512 bytes.
Tracked-On: #3138
Signed-off-by: Vijay Dhanraj <vijay.dhanraj@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Define a static mptable array and each VM could index its vmptable by
vm id, then mptable is not needed in vm configurations;
Tracked-On: #2291
Signed-off-by: Victor Sun <victor.sun@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
The CCS bit of PORTSC register should be set according to the mapped
native port connection status, use xdev->devices if equal NULL is not
enough due to devices only be clear in disable slot which can't reflect
the native connection state in some gaps.
Tracked-On: #3163
Signed-off-by: Xiaoguang Wu <xiaoguang.wu@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
There are many quirks in Intel xHCI implementation and it is very hard
to debug under Microsoft Windows OS, hence use ACRN xHCI extented
capabilities as the default setting.
Tracked-On: #3163
Signed-off-by: Xiaoguang Wu <xiaoguang.wu@intel.com>
Acked-by: Yu Wang <yu1.wang@intel.com>
Victor Sun