ACRN does not support platforms that do not have x2APIC mode of LAPIC
in hardware. With this patch, x2APIC is exposed to guests by default.
Extended Topology Leaf 0xb in cpuid returns x2APIC ID and topology
information to OS. This patch adds support to return guest topology
and guest x2APIC ID. Number of SMT siblings is returned as 0.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
This patch adds support for self-IPI virtualization when guest
uses vLAPIC in x2APIC mode.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
When guest switches from xAPIC mode to x2APIC mode of vLAPIC operation,
MSRs are used to access vLAPIC. This patch adds APICv support for
MSR accesses to vLAPIC. Switching from xAPIC to x2APIC is supported via
APIC BASE MSR. Other modifications like disabling and switching back to
xAPIC are not supported.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
This patch adds support to return APIC ID if guest uses vLAPIC
in x2APIC mode.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
This patch converts x2APIC MSR accesses to corresponding LAPIC MMIO offset to
utitlize vlapic_write/read APIs to virtualize LAPIC. Also adds support to inject
GP fault when read-only registers are attempted to be written to or vice versa.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
When guest uses vLAPIC in x2APIC mode, ICR write is a single MSR write.
Also, the destination field for device interrupts and IPIs should not be
handled in the same way as xAPIC mode. This patch adds support for x2APIC
mode operation of guest vLAPIC.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
x2APIC ID and LDR are in different format compared to xAPIC mode of operation.
This patch adds code to build ID and LDR when guest uses vLAPIC in x2APIC mode.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
Extending enable_msr_interception to accept mode as input. Mode specifies
if the API user wants ACRN to intercept on read-only or write-only or both
read and write or disable MSR interception altogether.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
Accessing x2APIC MSRs in xAPIC mode should result in GP exception according
to SDM section 10.12.2. Adding support to ACRN to inject GP into guests for
the same reason.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Xu Anthony <anthony.xu@intel.com>
In the C99 standard, the order of evaluation associated with multiple #,
multiple ## or a mix of # and ## preprocessor operator is unspecifie.
So it is unsafe to use multiple # or ## in a macro.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Normal world would also setup 511 GB gpa EPT mapping when initialize.
So we couldn't know which the world is from the gpa. But trusty is so
special for that we know where it would get a pml4_page or pdpt_page.
As a result, we could simpler this by just return the pml4_page or
pdpt_page to it when it needs.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Replace dynamic memory allocation in memory management with static memory allocation.
Since the static memory allocation can guarantee the allocation never failed, so
split_large_page and construct_pgentry don't need to return a errno any more.
Besides, the destroy_ept don't need to free page any more. Instead, it would memset
the eptp to 0 to make sure we can't walk this paging table again.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Add static paging table allocation API for EPT.
Note: must configure SOS/UOS_REAM_SIZE exactly as the platform.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Add static paging table allocation API for hypervisor.
Note: must configure PLATFORM_RAM_SIZE and PLATFORM_MMIO_SIZE exactly as the platform.
Rename RAM_START/RAM_SIZE to HV_RAM_START/HV_RAM_SIZE for HV.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
On MRB, there are some differences for SOS uart setting between
debug version and release version:
for debug version, ttyS0 is vuart, ttyS1 is 00:18.0, ttyS2 is 00:18.1,
ttyS3 is 00:18.3.
for release version, ttyS0 is 00:18.0, ttyS1 is 00:18.1, ttyS2 is
00:18.2, ttyS3 is 00:18.3. There is no vuart.
port: 00:18.0 is for bluetooth usage, ttyS0 can't be used as SOS console
for release build.
after change, debug build: vuart -->ttyS2 (SOS console);
release version: ttyS2; 00:18.2, it can be used as SOS console too.
for UEFI platform, also change acrn.conf console=ttyS0-->ttyS2
to matched with vart change.
Tracked-On: #1690
Signed-off-by: Minggui Cao <minggui.cao@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
It's possible that the up_count_spinlock is not release during
system enter S3. The case is like following:
BSP AP
stop_cpus cpu_dead
cpu_set_current_state
spinlock_abtain
up_count--
wait_for up_count == 1
enter S3
spinlock_release
Especially, considering the real spinlock release could be delayed
by cache.
Actually, the most content protected by up_count_spinlock is per
cpu data and could be changed without lock. Only left is up_count.
This patchset remove the up_count_spinlock and use atomic API
for up_count changing.
Tracked-On: #1691
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Eddie Dong <eddie.dong@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
add size check for other hypervisor console command;
they could be overflow for shell log buffer output.
Tracked-On: #1587
Signed-off-by: Minggui Cao <minggui.cao@intel.com>
Reviewed-by: Yin Fengwei <fengwei.yin@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
on KBL-NUC when input "vmexit" in hypervisor console,
the console or HV/SOS could be hung, the root cause is:
the log buffer is overflow for 8 CPU cores info.
to resolve the issue:
1. increase the shell log buffer size according to the
physical CPU max number
2. check the snprintf return value, if no buffer left,
just return.
Tracked-On: #1587
Signed-off-by: Minggui Cao <minggui.cao@intel.com>
Reviewed-by: Yin Fengwei <fengwei.yin@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
After enabling vPCI in hypervisor for vm0, UOS may not able be launched
successfully. Consider this scenario (take MSI-X for example):
- DM makes hypercall to hypervisor to do MSI-X remapping on behalf of
UOS guests.
- After the hypercall, VHM module in SOS kernel updates the physical
MSI-X table with the physical Message Data/Addr.
- These MMIO write requests are intercepted by hypervisor, which will
call ptdev_msix_remap() to do MSI-S remapping.
It may fail due to 2 possible reasons:
1) wrong target VM because:
hypervisor thinks it's a VM0 MSI-X device but they have been registered
as UOS guests through HC_SET_PTDEV_INTR_INFO hypercall.
2) wrong ptdev_msi_info->vmsi_data because:
The virtual MSI-X table is supposed to hold virtual Message data/addr
but the SOS VHM writes the physical ones to it.
This patch resolves these problems by ignoring the HC_VM_PCI_MSIX_REMAP
hypercall, so virtual and physical Message Data are the same from SOS'
perspective and it won't mess up the virtual PCI device in HV.
Also in HC_SET_PTDEV_INTR_INFO handler, vpci updates the target VM
when the PCI devices are assigned to different VMs.
The UOS' MSI/MSI-X remapping is triggered by hypervisor when SOS (either
DM or VHM) updates the Message Data/Addr.
Tracked-On: #1568
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
If PBA or other data structures reside in the same BAR with MSI-X,
devicemodel could emulate them and make hypercall SET_MEMORY_REGION
to setup EPT for vm0.
Hypervisor can not intercept the whole range of this BAR, but only
the minimum 4KB pages that cover the MSI-X Table.
Tracked-On: #1568
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Some PCI devices can not accept DWORD write to the MSI-X Table though
PCI Spec allows it:
PCI Spec 3.0 6.8.3.2 MSI-X Configuration: Specifically, software is
always permitted to fill in the Message Address and Message Upper Address
fields with a single QWORD write.
Tracked-On: #1568
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Fixed the following violations:
- Widening cast on complex integer expression
- Casting operation to a pointer
- Static procedure is not explicitly called in code analysed
Tracked-On: #1568
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
To sizing a 64-bit BAR, need to form the two 32-bit registers as a
64-bit words, before doing the calculation: inverting all bits and
incrementing by 1.
Tracked-On: #1568
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Remove the check to multi-function flag in PCI enumeration because when
this flag is set, Device IDs and Class codes, and this flag may vary
between functions.
We don't assume checking this bit in function 0 is safe enough.
Tracked-On: #1568
Signed-off-by: dongshen <dongsheng.x.zhang@intel.com>
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
For SOS BSP, we reuse native saved cs.limit
For UOS BSP, we set cs.limit in DM
For AP, we use initialized data from realmode_init_regs.
Tracked-On: #1231
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Acked-by: Eddie Dong <Eddie.dong@intel.com>
Save the native cs limit and reuse it for SOS cs limit. To remove
the hardcode of cs limit for guest in hypervisor.
Tracked-On: #1231
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Acked-by: Eddie Dong <Eddie.dong@intel.com>
--Config LOG_BUF_SIZE 256KB for per cpu
--Replace 'calloc' with static array for sbuf
--Rename 'alloc_earlylog_sbuf' to 'init_earlylog_sbuf'
--Remove deadcode sbuf_free
v2-->v3:
-- put the buffer into per_cpu data structure
v1-->v2:
-- add 'is_early_logbuf' in percpu data structure used for
check if need to do 'do_copy_earlylog'
Tracked-On: #861
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Reviewed-by: Yan, Like <like.yan@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
-to avoid information leaking from secure world to
normal world on platform affected by L1TF.
Tracked-On: #1672
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
- flush L1 cache before VM entry only on platform
affected by L1TF
- flush operation is configurable by below MACRO:
--CONFIG_L1D_FLUSH_VMENTRY_ENABLED
Tracked-On: #1672
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
- detect if current processor is affected by L1TF
- detect the presence of of "IA32_FLUSH_CMD(MSR 0x10B)
if processor is affected by L1TF.
Tracked-On: #1672
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
- add a function to detect CPU capabilities for security
related features, it will print warning messages if
current ucode can't support the latest security features.
Tracked-On: #1672
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Reviewed-by: Kevin Tian <kevin.tian@intel.com>
For partition mode, there is no vm0_domain created for iommu.
In current code, it will return error when try to remove a device
from vm0_domain, which casue failure of device assignment in
partition mode.
Tracked-On: #1680
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
The operands to shift operations (<<, >>) shall be unsigned integers.
v1 -> v2:
* replace 12U with CPU_PAGE_SHIFT when it is address shift case.
* replace 6UL with 0x6UL
Tracked-On: #861
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Replace list of iommu_domains with static array
Tracked-On: #861
Signed-off-by: Tw <wei.tan@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Predefine a dmar_drhd_rt table for all possible dmar unit
Tracked-On: #861
Signed-off-by: Tw <wei.tan@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
%x could be used to replace the %o print option.
%x could be used to replace the %p print option also.
Tracked-On: #1656
Signed-off-by: Yin Fengwei <fengwei.yin@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Below functions are deprecated, remove them from hypercall.h
hc_assert_irqline()/hc_deassert_irqline()/hc_pulse_irqline()
Tracked-On: #861
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Guest OS can use vLAPIC in x2APIC mode. Upon vlapic_reset, vLAPIC should
be set to xAPIC mode.
Tracked-On: #1626
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Instead of using the ACPI info template in the source tree, this patch requires
a board-specific ACPI info header to be created by the user and placed under
bsp/include. Currently we'll fall back to the current platform-specific template
if such board-specific info is not available.
A configuration symbol ACPI_ENFORCE_VALIDATED_ACPI_INFO is also added to enforce
the existance of board-specific ACPI info header. Default configurations can set
this symbol if the template does not work on a board.
Updates to the getting started guide will be updated accordingly after the
offline tools to generate such header is available.
v2 -> v3:
* Rephrase the ACPI-info-related messages.
v1 -> v2:
* The generated header should depend on .config so that whenever any
configuration changes, this header will be remade.
Tracked-On: #1520
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
Reviewed-by: Victor Sun <victor.sun@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
Currently silentoldconfig uses the default values in Kconfig when .config does
not exist or is for a different platform from the one specified in command line
options. This makes it impossible to leverage defconfigs when a simple 'make' is
invoked to build the hypervisor, as 'make' calls oldconfig instead of defconfig
for creating a configuration file.
With DEFCONFIG_LIST in Kconfig, now silentoldconfig is also able to know which
defconfig is appropriate given the specified board. This patch changes the
behavior of oldsilentconfig as follows.
1. When an old .config for the given board exists, the script updates the
.config per configurations specified from the command line. Instead of
manually resetting all symbols, the script now relies on Kconfiglib's
internal logic to automatically update invisible symbols after symbol
value changes.
2. When an old .config does not exist or is not for the specified board, the
defconfig for that board is used as base line and further tweaked per
configurations from the command line.
v3 -> v4:
* Rephrase comments in silentoldconfig.py.
v2 -> v3:
* Keep backward compatibility of silentoldconfig.py since vmcfg also uses the
script.
v1 -> v2:
* Include .config when it is edited manually so that new configurations
overwritten the old (in the outdated config.mk).
* Compare the board in .config with the one in the defconfig instead of what is
specified in the environment variable BOARD, as defconfig and BOARD may use
different names (e.g. aliases or different cases)
Tracked-On: #1588
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
The current defconfigs are BIOS-specific which makes it difficult to maintain
multiple defconfigs for boards running the same BIOS.
This patch re-organizes the defconfigs to be board-specific. A command line
option BOARD is introduced to specify a board on which the current build targets
at. The original PLATFORM is kept for backward compatibility which redirects to
apl-mrb and nuc6cayh for sbl and uefi, respectively.
The getting started guide is also updated accordingly.
v1 -> v2:
* Rewrite 'up2' to 'UP2'.
Tracked-On: #1588
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
Without any command under the .config target, config.mk will not be remade in
the same execution of make after oldconfig changes .config. This can result to
inconsistency between .config and the built image right after .config is
changed.
Tracked-On: #1588
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
It is faster to check the existance of a certain library by trying importing
that library, instead of invoking pip3 for a complete list of installed
libraries. Time of the check can be significantly reduced.
# time pip3 list
...
real 0m6.038s
user 0m0.652s
sys 0m0.036s
# time python3 -c "import kconfiglib"
real 0m0.037s
user 0m0.036s
sys 0m0.000s
Tracked-On: #1588
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
Add help messages to all visible symbols except those heap-related ones which
will be dropped eventually.
v2 -> v3
* Fix typos and misleading descriptions.
v1 -> v2
* Per kconfig language, help messages shall have a larger indent level than the
attributes.
* Rephrase for clarify.
Tracked-On: #1588
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
