For FuSa's case, we remove all dynamic memory allocation use in ACRN HV. Instead,
we use static memory allocation or embedded data structure. For pagetable page,
we prefer to use an index (hva for MMU, gpa for EPT) to get a page from a special
page pool. The special page pool should be big enougn for each possible index.
This is not a big problem when we don't support 64 bits MMIO. Without 64 bits MMIO
support, we could use the index to search addrss not larger than DRAM_SIZE + 4G.
However, if ACRN plan to support 64 bits MMIO in SOS, we could not use the static
memory alocation any more. This is because there's a very huge hole between the
top DRAM address and the bottom 64 bits MMIO address. We could not reserve such
many pages for pagetable mapping as the CPU physical address bits may very large.
This patch will use dynamic page allocation for pagetable mapping. We also need
reserve a big enough page pool at first. For HV MMU, we don't use 4K granularity
page table mapping, we need reserve PML4, PDPT and PD pages according the maximum
physical address space (PPT va and pa are identical mapping); For each VM EPT,
we reserve PML4, PDPT and PD pages according to the maximum physical address space
too, (the EPT address sapce can't beyond the physical address space), and we reserve
PT pages by real use cases of DRAM, low MMIO and high MMIO.
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Tracked-On: #5788
Rename hv_access_memory_region_update to ppt_clear_user_bit to
verb + object style.
Tracked-On: #5330
Signed-off-by: Li Fei1 <fei1.li@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Temporarily remove NX bit of PTCM binary in pagetable during pSRAM
initialization:
1.added a function ppt_set_nx_bit to temporarily remove/restore the NX bit of
a given area in pagetable.
2.Temporarily remove NX bit of PTCM binary during pSRAM initialization to make
PTCM codes executable.
3. TODO: We may use SMP call to flush TLB and do pSRAM initilization on APs.
Tracked-On: #5330
Signed-off-by: Qian Wang <qian1.wang@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
PCI devices with 64-bit MMIO BARs and requiring large MMIO space
can be assigned with physical address range at the very high end of
platform supported physical address space.
This patch uses the board info for 64-bit MMIO window as programmed
by BIOS and constructs 1G page tables for the same.
As ACRN uses identity mapping from Linear to Physical address space
physical addresses upto 48 bit or 256TB can be supported.
Tracked-On: #4586
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Currently ept_pages_info[] is initialized with first element only that force
VM of id 0 using SOS EPT pages. This is incorrect for logical partition and
hybrid scenario. Considering SOS_RAM_SIZE and UOS_RAM_SIZE are configured
separately, we should use different ept pages accordingly.
So, the PRE_VM_NUM/SOS_VM_NUM and MAX_POST_VM_NUM macros are introduced to
resolve this issue. The macros would be generated by acrn-config tool when
user configure ACRN for their specific scenario.
One more thing, that when UOS_RAM_SIZE is less then 2GB, the EPT address
range should be (4G + PLATFORM_HI_MMIO_SIZE).
Tracked-On: #4458
Signed-off-by: Victor Sun <victor.sun@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
- change variable name from hpa to hva because in this function we are
dealing with hva, not hpa.
- can get the address of ld_text_end by directly referring to this symbol,
because relative addressing yields the correct hva, not the hva before
relocation.
Tracked-On: #4441
Signed-off-by: Zide Chen <zide.chen@intel.com>
Now the e820 structure store ACRN HV memory layout, not the physical memory layout.
Rename e820 to hv_hv_e820 to show this explicitly.
Tracked-On: #4007
Signed-off-by: Li Fei1 <fei1.li@intel.com>
change the input parameter from vcpu to eptp in order to let this api
more generic, no need to care normal world or secure world.
Tracked-On: #1842
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
EPT tables are shared by MMU and IOMMU.
Some IOMMUs don't support page-walk coherency, the cpu cache of EPT entires
should be flushed to memory after modifications, so that the modifications
are visible to the IOMMUs.
This patch adds a new interface to flush the cache of modified EPT entires.
There are different implementations for EPT/PPT entries:
- For PPT, there is no need to flush the cpu cache after update.
- For EPT, need to call iommu_flush_cache to make the modifications visible
to IOMMUs.
Tracked-On: #3607
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Reviewed-by: Anthony Xu <anthony.xu@intel.com>
1) Using printf to warn if platform ram size configuration is wrong.
2) Using printf to warn if the platform is not supported by ACRN hypervisor.
Tracked-On: #2987
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
flush_address_space is used to flush address space by clflushopt instruction.
Signed-off-by: Jack Ren <jack.ren@intel.com>
Signed-off-by: Yuan Liu <yuan1.liu@intel.com>
Reviewed-by: Li, Fei1 <fei1.li@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Host ACPI parsing is needed during initialization only, not in run time.
Hence we don't need to clear U flag for memory in reserved or ACPI type
E820 entries.
- move enable_smep() and enable_smap() to the end of init_pcpu_post(),
so stac()/clac() can be removed from any init code before this point.
- call init_seed() before init_pcpu_post(), and rmeove stac()/clac() from
init_seed().
Tracked-On: #3194
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Previously we use Kconfig of DMAR_PARSE_ENABLED to choose pre-defined DMAR info
or parse it at runtime, at the same time we use MACRO of CONFIG_CONSTANT_ACPI
to decide whether parse PM related ACPI info at runtime. This looks redundant
so use a unified ACPI_PARSE_ENABLED Kconfig to replace them.
Tracked-On: #3107
Signed-off-by: Victor Sun <victor.sun@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
This patch adds prefix 'p' before 'cpu' to physical cpu related functions.
And there is no code logic change.
Tracked-On: #2991
Signed-off-by: Kaige Fu <kaige.fu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Currently vpid is not released in reset_vcpu() hence the vpid resource
could be exhausted easily if guests are re-launched.
This patch assigns vpid according to the fixed mapping of runtime vm_id
and vcpu_id to guarantee the uniqueness of vpid.
Tracked-On: #2700
Signed-off-by: Zide Chen <zide.chen@intel.com>
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
The init page tables installed in either cpu_primary.S or trampoline.S
are 1:1 mapping and won't be changed in the future.
The 'actual' hypervisor page table installed in enable_paging() is 1:1
mapping currently but it could be changed in the future. Both hva2hpa() and
hpa2hva() are implemented based on these page tables and can't be used
when the init page tables take effect.
This patch does the following cleanup:
- remove all hva2hpa()/hpa2hva() before calling enable_paging()
- get_hv_image_base() returns HVA, not HPA. So add hva2hpa() for all cases
that are called afte enable_paging().
Tracked-On: #2700
Signed-off-by: Zide Chen <zide.chen@intel.com>
Acked-by: Eddie Dong <Eddie.dong@intel.com>
-- add header file ld_sym.h in include/arch/x86/boot/
-- move 'ld_bss_start/end' from cpu.h to ld_sym.h,
avoid reverse dependency
-- move 'ld_text_end' from mmu.h to ld_sym.h
-- move 'ld_trampoline_load/start/end' to ld_sym.h
Tracked-On: #1842
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
when relocation is enabled, the calcuation of HV size in init_paging
routine is not right. As a result, it doesn't enable NX bit for many
pages which don't belong to hypervisor text section.
Tracked-On: #2349
Signed-off-by: Chaohong guo <chaohong.guo@intel.com>
Reviewed-by: Zide Chen <zide.chen@intel.com>
On platforms that put ACPI RSDT in "Reserved" type memory, ACRN
fails to parse RSDP and therefore ACPI Tables. This patch adds
"Reserved" memory regions in E820 as Supervisor pages.
Tracked-On: #2320
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Fix violations that attempt to change parameter
passed by value.
Tracked-On: #2056
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
- bugfix:the actual 'size' of memory region that
to be updated is incorrect.
- replace CONFIG_UEFI_STUB with DMAR_PARSE_ENABLED
when update memory pages for ACPI_RECLAIM region,
as DMAR_PARSE_ENABLED may be enabled on non-EFI
platform.
V2 update:
wrap roundup to 2M and rounddown to 2M inline
functions.
Tracked-On: #2056
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Try to minimize the C code in inline assembly function. Now only
construct data structure and return a value is permitted.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
1) Separate ASM and C code inserts as pure procedures with a defined interface.
2) Fix procedure has more than one exit point.
Tracked-On: #861
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
move cpu caps related functions like capability init/detect/check
in cpu.c & mmu.c into a new file cpu_caps.c
Changes to be committed:
modified: developer-guides/hld/hv-memmgt.rst
modified: ../hypervisor/Makefile
modified: ../hypervisor/arch/x86/cpu.c
new file: ../hypervisor/arch/x86/cpu_caps.c
modified: ../hypervisor/arch/x86/mmu.c
modified: ../hypervisor/arch/x86/vmx_asm.S
modified: ../hypervisor/include/arch/x86/cpu.h
new file: ../hypervisor/include/arch/x86/cpu_caps.h
modified: ../hypervisor/include/arch/x86/guest/vm.h
modified: ../hypervisor/include/arch/x86/mmu.h
modified: ../hypervisor/include/arch/x86/vmcs.h
Tracked-On: #1842
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
With SMAP enabled, hypervisor can't access pages that
owned by guest(either SOS or UOS), and an override is
is provided: stac()/clac() to enable/disable access to
guest's memory pages.
Pre-conditon:
Mark hypervisor owned pages as supervisor mode (U/S = 0),
and set all othter memory pages as user mode (U/S = 1).
Tracked-On: #2056
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
- enable NX feature in hypervisor:
1. Set 'XD' bit for all pages, including pages for guests
when initialize MMU tables in hypervisor.
2. remove 'XD' bit for pages that contain hypervisor instructions.
3. enable MSR EFER.NXE,which will enable page access restriction by
preventing instruction fetches form pages with XD bit set.
- remove "-Wl -z noexecstack" GCC flag option in hypervisor
Makefile as it would not affect stack attribute in hyervisor,
which setup stack itself, instead of by loader.
Tracked-On: #1122
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
1. change its APIs as more indepentent, and modify the callers' code
2. limit its global variables as static, and return const to the callers
3. remove unused code in "CONFIG_CMA"
Tracked-On: #1842
Signed-off-by: Minggui Cao <minggui.cao@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
1. e820 is modulized as separated files.
2. move boot_regs into multiboot.h as it's related with
multiboot info header
Tracked-On: #1842
Signed-off-by: Minggui Cao <minggui.cao@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
IEC 61508,ISO 26262 standards highly recommend single-exit rule.
Reduce the count of the "return entries".
Fix the violations which is comply with the cases list below:
1.Function has 2 return entries.
2.The first return entry is used to return the error code of
checking variable whether is valid.
Fix the violations in "if else" format.
Tracked-On: #861
Signed-off-by: Huihuang Shi <huihuang.shi@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Now it's using the memory reserve by hypervisor. So there's not
necessaty to map or ummap this region from SOS.
Tracked-On: #1942
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
The previous would reserve memory for trusty in SOS kernel. Howerer,
there would no available 16 MB continue memory any more after a long time.
This result in allocating memory for trusty failed. This patch will reserve
memory for trusty in ACRN hypervisor in which case the memory allocation
for trusty will never fail.
Tracked-On: #1942
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
now the size of sanitized_page is 32KB,
revise it to 4KB.
Tracked-On: #1842
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
MISRA-C requires that the controlling expression of an if statement or
an iteration-statement shall be Boolean type.
v1 -> v2:
* update the U suffix of constant to let the type of operands
to bit operations (&) is the same.
* update the type of 'i' from uint64_t to uint16_t in
'profiling_set_control' since it is limited by phys_cpu_num,
which is uint16_t.
Tracked-On: #861
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
For data structure types "struct vm", its name is identical
with variable name in the same scope. This is a MISRA C violation.
Naming convention rule:If the data structure type is used by multi
modules, its corresponding logic resource is exposed to external
components (such as SOS, UOS), and its name meaning is simplistic
(such as vcpu, vm), its name needs prefix "acrn_".
The following udpates are made:
struct vm *vm-->struct acrn_vm *vm
Tracked-On: #861
Signed-off-by: Xiangyang Wu <xiangyang.wu@linux.intel.com>
For data structure types "struct vcpu", its name is identical
with variable name in the same scope. This is a MISRA C violation.
Naming convention rule:If the data structure type is used by multi
modules, its corresponding logic resource is exposed to external
components (such as SOS, UOS), and its name meaning is simplistic
(such as vcpu, vm), its name needs prefix "acrn_".
The following udpates are made:
struct vcpu *vcpu-->struct acrn_vcpu *vcpu
Tracked-On: #861
Signed-off-by: Xiangyang Wu <xiangyang.wu@linux.intel.com>
Unmap the trusty EPT mapping when destroy_secure_world. The reason is
now the normal world and trusty will share the PD and PT page. Before
we add the mapping for each world, we should unmap it from another world.
Besides, fix a minor bug for condition priority.
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 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>
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>
Fix violations whose parameter can be read-only.
This patch only fix the parameter whose name is vcpu.
Tracked-On: #861
Signed-off-by: Huihuang Shi <huihuang.shi@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
In the current hypervisor design, when HPA is not
found for the specified gpa by calling gpa2hpa or
local_gpa2hpa, 0 will be returned as a error code,
but 0 may be a valid HPA for vm0; error checking
is missed when invoking gpa2hpa or local_gpa2hpa;
when invoking lookup_address, the caller guarantees
that parameter pointer pml4_page and pointer pg_size
is not NULL.
If local_gpa2hpa/gpa2hpa returns a invalid HPA,
it means that this function fails to find the
HPA of the specified gpa of vm. If local_gpa2hpa/gpa2hpa
return value is a valid HPA, it means that this
function have found the HPA of the specified gpa of vm.
Each valid vm's EPTP is initialized during vm creating,
vm's EPTP is valid until this vm is destroyed. So the caller
can guarantee parameter pointer pml4_page is not NULL.
The caller uses a temporary variable to store page size.
So the caller can guarantee parameter pointer pg_size
is not NULL.
In this patch, define a invalid HPA for gpa2hpa and
local_gpa2hpa;add some error checking when invoking
local_gpa2hpa/gpa2hpa;add precondition for lookup_address
function and remove redundant error checking.
V1-->V2:
Define INVALID_HPA as a invalid HPA for gpa2hpa
and local_gpa2hpa;
Updated related error checking when invoking
gpa2hpa or local_gpa2hpa;
V2-->V3:
Add some debug information if specified gpa2hpa
mapping doesn't exit and ept_mr_del is called;
Update INVALID_HPA definition easier to be reviewed.
V3-->V4:
Add vm->id and gpa into pr_error;
Add precondition to ept_mr_del to cover [gpa,gpa+size)
unmapping case.
V4-->V5:
Update comments;
Update pr_error message.
Tracked-On: #1258
Signed-off-by: Xiangyang Wu <xiangyang.wu@linux.intel.com>
Reviewed-by: Li, Fei1 <fei1.li@intel.com>
There are chances that names with leading underscore declared by
developers are conflict with the ones reserved for the compiler.
What this patch does:
- rename these functions/variables/macros starting with
underscore to avoid such unintentational mistakes.
- remove gpr.h without any contents
Tracked-On: #861
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
