Add comma to the last member of ivshmem pci devices.
If the last element ends without comma, the clang-format would attach
the brackets to the first and last lines.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add a xslt file "misc_cfg.h.xsl". This file is used to
generate misc_cfg.h which is used by hypervisor.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add a xslt file "ivshemem_cfg.h.xsl". This file is used to
generate ivshemem_cfg.h which is used by hypervisor.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add a xslt file "pt_intx.c.xsl". This file is used to
generate pt_intx.c which is used by hypervisor.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add a xslt file "vm_configurations.h.xsl". This file is used to
generate vm_configurations.h which is used by hypervisor.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add a xslt file "vm_configurations.c.xsl". This file is used to
generate vm_configurations.c which is used by hypervisor.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
This file allocates the resource of pio base.
The available pio base is in ['0x3F8', '0x2F8', '0x3E8', '0x2E8'] and it
cannot be used by native device.
When any of sos legacy vuarts are enabled, assign a pio base to legancy
vuarts' base.
The allocator follows rules:
1. An SOS legacy vuart only support PIO vuart.
2. To assign a pio base for sos legacy vuart 0:
- If the hv/DEBUG_OPTIONS/SERIAL_CONSOLE is one of
[ttys0, ttys1, ttys2, ttys3] and it's a pio vuart in the native
environment, the pio base of SOS legacy vuart 0 would be the same as
native one.
- If the hv/DEBUG_OPTIONS/SERIAL_CONSOLE is not one of
[ttys0, ttys1, ttys2, ttys3], assigns a pio base to SOS legacy vuart 0
from avilable pio base.
- If the hv/DEBUG_OPTIONS/SERIAL_CONSOLE is not one of
[ttys0, ttys1, ttys2, ttys3] but a pio vuart, will assigns a pio
base to SOS legacy vuart 0 from avilable pio base and raise a
warning to user.
3. To assign a pio base for sos legacy vuart 1:
- Assigns a pio base to SOS legacy vuart 1 from avilable pio base.
- If all the available pio bases list is empty (which means all are
in used by native), it assigns one of the pio base to SOS legacy
vuart 1 anyway, but raise a warning to user.
4. Assigned pio bases must be unique.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Create an intx.py which is a static allocator to allocate the irq
resources. The available irq list is based on the native irqs which
are in range [0, 15] and are not used by native os.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
If a cpu_affinity node of SOS is not present in the scenario.xml,
assign the native cpus which are not assigned to pre-launched vm to
SOS vm.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Refine the "append_node" which can add new node with an attribute and
return the appended node.
The method "get_node" finds the xpath value and return it if there is an
unique node exists, otherwise it returns None.
It is used to get an xpath element node or can determine the xpath existence.
The "get_text" is replaced with "get_node". The only get_text in
hv_ram.py is modified accordingly.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Add lib.xsl under config_tools/xforms.
This lib.xsl contains the variables and customized functions for
xslt transformation.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Create lib.py which contains the common methods that are shared by static
allocators under misc/config_tools/static_allocators.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
Create a file which contains user-defined errors for config-tools.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
<guest_flag/> may be treated as either empty string or null in xslt
transformation and xsd schema validation. Replace it with:
<guest_flag></guest_flag>
to avoid the undefined behavior.
The duplicate guest_flag are removed.
Tracked-On: #5980
Signed-off-by: Yang,Yu-chu <yu-chu.yang@intel.com>
Reviewed-by: Junjie Mao <junjie.mao@intel.com>
This temporary release is based on the following commit:
commit 83b4731bab0423e3f9ffccfb37a882090be2f44a
Author: Tao Yuhong <yuhong.tao@intel.com>
Date: Wed Apr 14 07:43:11 2021 -0400
Change 64-bit MMIO BAR window to 256G-512G
DM maps 64-bit mmio BARs of vdev into 4G-5G, for post-launched VMs. At native
platform, 64-bit MMIO BARs which have 39-bit address, are always mapped into
256G-512G address space.
DM will change the address window of 64-bit vdev BARs of post-launched VMs to
256G-512G. That ask OVMF to do the same change, to boot from passthrough SATA/MVME
disks, which have 64-bit MMIO BAR.
Tracked-On: #5913
Signed-off-by: Peter Fang <peter.fang@intel.com>
ACRN didn't support dynamic memory allocation. SO it would reserve
a big page pool and use the GPA as index to get a page to do EPT
mapping. In order to save memory, we put high MMIO windows to [4G, 5G].
AFter we support dynamic page allocation for EPT mapping, we move
high MMIO windows to where it used to ([256G, 512G]) for 39 bits physical
address), we could move high memory to where is used to too ([4G, 4G +
size]).
Tracked-On: #5913
Signed-off-by: Tao Yuhong <yuhong.tao@intel.com>
DM maps 64-bit mmio BARs of vdev into 4G-5G, for post-launched VMs. At native
platform, 64-bit MMIO BARs which have 39-bit address, are always mapped into
256G-512G address space.
Change PCI_EMUL_MEMBASE64 to 256G, change PCI_EMUL_MEMLIMIT64 to 512G. So that
the 64-bit vdev BARs of post-launched VMs have same address space with native
platform.
Tracked-On: #5913
Signed-off-by: Tao Yuhong <yuhong.tao@intel.com>
Acked-by: Wang, Yu1 <yu1.wang@intel.com>
clang-format is now used as part of the config tools creating c files
based on the XML configuration
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
Add a note to the "Device Model Parameters" document to emphasize
the need to use the '--windows' parameter to use Windows-as-a-Guest
(WaaG), else Windows will not recognize the virtual disk it has
been assigned.
Tracked-On: #5962
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
We should only map [low32_max_ram, 4G) MMIO region as UC attribute,
not map [low32_max_ram, low32_max_ram + 4G) region as UC attribute.
Otherwise, the HV will complain [4G, low32_max_ram + 4G) region has
already mapped.
Tracked-On: #5830
Signed-off-by: Li Fei1 <fei1.li@intel.com>
This patch fixes the 'uart=bdf@XXX' mechanism for the PCI serial
port devices which bar0 is not MMIO.
Tracked-On: #5968
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
Signed-off-by: Li Fei <fei1.li@intel.com>
GPA of software SRAM is available only after
build_vrtct() function is called and the return
value of it is valid(Not NULL).
This patch fix bug in create_and_inject_vrtct()
function which violates above pre-condition when
calling get_software_sram_base_gpa().
Tracked-On: #5973
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Update known-issues pattern for PDF processing to also work with updated
xelatex tools from Ubuntu 20.04
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
PRs #5945 and #5949 introduced fixes to the doc building process to
support PDF generation of the documentation set. This PR refines the
doc build process, cleaning up the Makefile, adding display of tool
version information, and updates the doc building documentation to
include additional dependencies needed for building the PDF and
instructions for how to build the PDF. The latexpdf make target is
provided to just run the latex and PDF producing process that depends on
the HTML artifacts from a make html run. A new make pdf target is
provided that combines the two steps into one.
A new know-issues pattern file is added that verifies the expected
output from the latexpdf process is returned, as it can't be completely
eliminated without losing potential error messages that need to be
resolved.
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
add 2 * max (IVSHMEM_SHM_SIZE, 2M) in HV_RAM_SIZE calculation to
avoid ram overflow caused by additional memory usage for shared
memory alignment.
Tracked-On: #5955
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Reviewed-by: Victor Sun <victor.sun@intel.com>
Both Windows guest and Linux guest use the MSR MSR_IA32_CSTAR, while
Linux uses it rarely. Now vcpu context switch doesn't save/restore it.
Windows detects the change of the MSR and rises a exception.
Do the save/resotre MSR_IA32_CSTAR during context switch.
Tracked-On: #5899
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
TLFS spec defines that when a VM is created, the value of
HV_X64_MSR_TIME_REF_COUNT is set to zero. Now tsc_offset is not
supported properly, so guest get a drifted reference time.
This patch implements tsc_offset. tsc_scale and tsc_offset
are calculated when a VM is launched and are saved in
struct acrn_hyperv of struct acrn_vm.
Tracked-On: #5956
Signed-off-by: Jian Jun Chen <jian.jun.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
TLFS spec defines that HV_X64_MSR_VP_INDEX and HV_X64_MSR_TIME_REF_COUNT
are read-only MSRs. Any attempt to write to them results in a #GP fault.
Fix the issue by returning error in handler hyperv_wrmsr() of MSRs
HV_X64_MSR_VP_INDEX/HV_X64_MSR_TIME_REF_COUNT emulation.
Tracked-On: #5956
Signed-off-by: Jian Jun Chen <jian.jun.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
TLFS spec defines different hypercall ABIs for X86 and x64. Currently
x64 hypercall interface is not supported well.
Setup the hypercall interface page according to the vcpu mode.
Tracked-On: #5956
Signed-off-by: Jian Jun Chen <jian.jun.chen@intel.com>
Signed-off-by: Shuo A Liu <shuo.a.liu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
These two MACROs shall be wrapped as a single
value respectively, hence brackets should be used.
Tracked-On: #5951
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Update missing captions on figures to remove remaining broken references
during latexpdf building. Also, require doing a "make html" before
doing a "make latexpdf" to build all the artifacts needed for running
the latexpdf build. (We might change that later if needed.)
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
In order to support platform (such as Ander Lake) which physical address width
bits is 46, the current code need to reserve 2^16 PD page ((2^46) / (2^30)).
This is a complete waste of memory.
This patch would reserve PD page by three parts:
1. DRAM - may take PD_PAGE_NUM(CONFIG_PLATFORM_RAM_SIZE) PD pages at most;
2. low MMIO - may take PD_PAGE_NUM(MEM_1G << 2U) PD pages at most;
3. high MMIO - may takes (CONFIG_MAX_PCI_DEV_NUM * 6U) PD pages (may plus
PDPT entries if its size is larger than 1GB ) at most for:
(a) MMIO BAR size must be a power of 2 from 16 bytes;
(b) MMIO BAR base address must be power of two in size and are aligned with
its size.
Tracked-On: #5929
Signed-off-by: Li Fei1 <fei1.li@intel.com>
The platform which physical-address width over 39 bits must support
1GB large page (Both MMU and VMX sides ). This could save lots of
page table pages for EPT MMIO mapping.
Tracked-On: #5929
Signed-off-by: Li Fei1 <fei1.li@intel.com>
This patch tweaks the settings in doc/conf.py to allow formatting the
documentation to a PDF file by Sphinx. The changes include:
- Use `xelatex` rather than the default `pdflatex` as the LaTeX engine, as
`pdflatex` is not that good at formatting non-ascii characters out of
the box.
- Use DejaVu fonts (which are available in common Linux distributions) in
the generated PDF.
- Restrict the depths of the table of contents to 3.
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
No one uses get_mem_range_info to get the top/bottom/size of the physical memory.
We could get these informations by e820 table easily.
Tracked-On: #5830
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Acked-by: eddie Dong <eddie.dong@intel.com>
We used get_mem_range_info to get the top memory address and then use this address
as the high 64 bits max memory address of SOS. This assumes the platform must have
high memory space.
This patch removes the assumption. It will set high 64 bits max memory address of
SOS to 4G by default (Which means there's no 64 bits high memory), then update
the high 64 bits max memory address if the SOS really has high memory space.
Tracked-On: #5830
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Acked-by: eddie Dong <eddie.dong@intel.com>
SOS's memory size could be calculated by its vE820 Tables easily.
Tracked-On: #5830
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Acked-by: eddie Dong <eddie.dong@intel.com>
We used get_mem_range_info to get the top memory address and then use this address
as the high 64 bits max memory address. This assumes the platform must have high
memory space.
This patch calculates the high 64 bits max memory address according the e820 tables
and removes the assumption "The platform must have high memory space" by map the
low RAM region and high RAM region separately.
Tracked-On: #5830
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Acked-by: eddie Dong <eddie.dong@intel.com>
Now BSP may launch VMs before APs have not done its initilization,
for example, sched_control for per-cpu. However, when we initilize
the vcpu thread data, it will access the object (scheduler) of the
sched_control of APs. As a result, it will trigger the PF.
This patch would waits each physical has done its initilization before
to continue to execute.
Tracked-On: #5929
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Using the MFENCE to make sure trampoline code
has been updated (clflush) into memory beforing start APs.
Tracked-On: #5929
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Make the description of the "Logial Partitioning" scenario more
generic than what is shown on the figure. This also helps as the
current examples of that scenario in the code base do not use
Safety or RTVM at the moment (as shown on the picture).
Tracked-On: #5903
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
Use MFENCE to strengthen the fast string operations execute order to ensure
all trampoline code was updated before flush it into the memory.
Tracked-On: #5929
Signed-off-by: Li Fei1 <fei1.li@intel.com>
fail to create kata vm type in industry scenario due to
the default vm id value is 1. Meanwhile set the max user
vm to 7 in tgl-rvp industry xml.
Tracked-On: #5932
Signed-off-by: lirui34 <ruix.li@intel.com>
There is no audio device in the default ehl-crb-b.xml, so remove
passthru audio devices from launch xmls on ehl-crb-b.
Tracked-On: #5925
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Reviewed-by: Victor Sun <victor.sun@intel.com>
remove UOS_RAM_SIZE and SOS_RAM_SIZE in scenario config since these
two config elements are useless.
Tracked-On: #5927
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Reviewed-by: Victor Sun <victor.sun@intel.com>
update HV_RAM_SIZE calculation algorithm to 20MB + VM number*
16MB, which consists of text segment rodata(2MB), bss data(about
1MB), bss.ppt_pages(2.4MB), bss.ctx_tables(6MB), bss.vm_array(
3.2MB), bss.ivshmem_base(2MB+1.8MB for alignment) and
bss.post_uos_sworld_memory(16MB*post-launched VM number).
Tracked-On: #5927
Signed-off-by: Shuang Zheng <shuang.zheng@intel.com>
Reviewed-by: Victor Sun <victor.sun@intel.com>
