Versions of kconfiglib lower than 10.2 cannot build the documentation
correctly. The minimum version required is >=10.2. This patch updates
the 'requirements.txt' accordingly.
Tracked-On: #2429
Signed-off-by: Geoffroy Van Cutsem <geoffroy.vancutsem@intel.com>
This example script launch a VM with lapic passthrough for realtime scenarios.
Tracked-On: #2351
Signed-off-by: Yan, Like <like.yan@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
New option "--lapic_pt" added to create VM with local apic passthrough, for
realtime scenarios.
When the option is set, a VM is created with LAPIC_PASSTHROUGH.
The option is not set by default.
Tracked-On: #2351
Signed-off-by: Yan, Like <like.yan@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
- don't generate io apic entry for vm with lapic pt.
Tracked-On: #2351
Signed-off-by: Yan, Like <like.yan@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
For VM with local apic pt for realtime scenatios, we support virtio device with PMD backend.
But we still need to inject MSI to notify the front-end, to avoid changing the front-end drivers.
Since the lapic is passed through, irq injection to vlapic won't work.
This commit fix it by sending IPI with vector need to inject.
Tracked-On: #2351
Signed-off-by: Yan, Like <like.yan@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Since lapic is passed through, MSI/MSIx should be configured with the
vector allocated within guest OS, instead of physical vector allocated
in hypervisor.
Tracked-On: #2351
Signed-off-by: Yan, Like <like.yan@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
This commit extend lapic pass-through for DM launched VM, generally for hard RT scenarios.
Similar to the partition mode, the vlapic is working under the xapic mode at first, only
when x2apic mode is enabled, lapic is passed through, because the physical LAPICs are
under x2apic mode.
Main changes includes:
- add is_lapic_pt() to check if a vm is created with lapic pt or not, to combine
codes of partition mode and DM launched vm with lapic passthrough, including:
- reuse the irq delievery function and rename it to dispatch_interrupt_lapic_pt();
- reuse switch_apicv_mode_x2apic();
- reuse ICR handling codes to avoid malicious IPI;
- intercept ICR/APICID/LDR msr access when lapic_pt;
- for vm with lapic passthrough, irq is always disabled under root mode.
Tracked-On: #2351
Signed-off-by: Yan, Like <like.yan@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
- vlapic_pending_intr() is called only for case no
APICv/VID(Virtual Interrupt Delivery) feature support,
in this case, there is no need to test VID case.
- remove 'apicv_pending_intr()' function.
Tracked-On: #2427
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
As we enable Interrupt Remapping, bit positions in IOAPIC RTEs
have a different syntax for programming. ACRN should handle original
format for vIOAPIC as well IR representation for physical IOAPIC.
This patch adds bit granularity IOAPIC RTE.
Tracked-On: #2407
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
As we enable Interrupt Remapping, bit positions in MSI address and
data registers have a different syntax for programming. This patch adds
bit granularity for MSI address and data structs.
Tracked-On: #2407
Signed-off-by: Sainath Grandhi <sainath.grandhi@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>
The sos_vm config under arch/x86/configs/$(CONFIG_BOARD) folder should be
customer specific configuration, leave it back to customer.
When hypervisor start building, it will check whether BOARD specific sos_vm.h
exist in configs/$(CONFIG_BOARD) folder. If exist, then include this header;
if not, then include default sos_vm config header under configs/default folder.
Tracked-On: #2291
Signed-off-by: Victor Sun <victor.sun@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
Add a API to sanitize vm_configs[] array, in this API, we will do sanity
check for VM configurations and refill some fields in vm_configs[]. If
sanity check is failed, the HV would reject to boot.
About pcpu_bitmap of SOS_VM:
We should not hard code pcpu_bitmap for SOS_VM, this is not convenient for
unconfigured boards. The pcpu_bitmap follows a very simple rule:
All physical CPUs except ocuppied by Pre-launched VMs are all belong to
SOS_VM. In this way, the pcpu_bitmap of a SOS_VM is decided by pcpu_bitmap
status in PRE_LAUNCHED_VMs.
To get the correct pcpu_bitmap of SOS_VM, We need to setup another rule, that
the vm_configs[] array should follow the order of PRE_LAUNCHED_VM - SOS_VM
strictly.
With this patch enabled, the pcpu_bitmap field of sos vm config is not needed
to configure;
Tracked-On: #2291
Signed-off-by: Victor Sun <victor.sun@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
after compile, the compiled code could change rsp, so use pure asm code
to avoid such problem which will cause schedule switch failure.
Tracked-On: #2410
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
'vector' should be no greater than 0xff,else
'eoi_exit_bitmap[]' will overflow.
Tracked-On: #1252
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
below fields are defined but not used.
- sync, pause_cnt & dbg_req_state.
Tracked-On: #861
Signed-off-by: Yonghua Huang <yonghua.huang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
The array indexx of "vm_list" may be out of bound.
Updated the size of "vm_list"
Bug: ACRN-2544
Tracked-On: #2385
Signed-off-by: Manisha Chinthapally <manisha.chinthapally@intel.com>
Previously I/O emulation completion mode mode was configured in Kconfig:
config IOREQ_NOTIFICATION
bool "Notification mode"
help
When I/O request is completed, SOS will mark the completion status and
notify hypervisor via hypercall. Hypervisor will finish the post work
when notification is received.
config IOREQ_POLLING
bool "Polling mode"
help
When I/O request is completed, SOS will only mark completion status
without notifying hypervisor. Hypervisor will poll the completion
status and finish the post work.
Now move this configuration to guest_flags of acrn_vm_config struct.
if ((vm_config->guest_flags & IO_COMPLETION_POLLING) != 0U) {
I/O with polling;
} else {
I/O with notification;
}
Tracked-On: #2291
Signed-off-by: Victor Sun <victor.sun@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
There is a missing dependency between lib and bin in Makefile.
BTW, this issue is triggered by concurrent build.
Tracked-On: #2370
Signed-off-by: Tw <wei.tan@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
Start a new FAQ document with questions we're often being asked on
support channels. We'll expand this list over time, and as the number
of questions grow, we'll organize them as appropriate.
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
Add a tutorial explaining how to set up a demo using AGL as UOSes as
used in a CES automotive demo.
Signed-off-by: David B. Kinder <david.b.kinder@intel.com>
In write_trampoline_stack_sym(), when do relocation for pcpu's stack,
it wrongly adds an offset. As a result, during booting, when non-BSP
CPU (APs) tries to run trampoline code once they are waken up by BSP
via SIPI, system hangs.
Tracked-On: #2349
Signed-off-by: Chaohong guo <chaohong.guo@intel.com>
Reviewed-By: Yin Fengwei <fengwei.yin@intel.com>
Reviewed-by: Zide Chen <zide.chen@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>
In init_paging(), when modify page table, it does a round up to 2M on
the starting address of HV memory. If the address is not 2M aligned,
memory overwrite occurs, or memory attribute(cache, u/s) may be not
right. Besides, trusty memory (uos_sworld_memory) in .bss section is 2M
aligned during compiling, after enabling HV memroy 2M alignment, when
relocation is enabled, EFI stub or other loaders just needs to find a
2M-aligned starting addr for HV memory.
Tracked-On: #2349
Signed-off-by: Chaohong guo <chaohong.guo@intel.com>
Reviewed-by: Zide Chen <zide.chen@intel.com>
efi_ctx is passed by EFI stub to hypervisor. The memory was allocated
right after HV binary and marked as Efireserved. But HV is doing a 2MB
alignment in init_paging() and might overwrite the efi_ctx struct or
change the page table attribute. Now, EFI STUB uses Efiloaderdata type
memory and the memory can be re-use by hypervisor/sos after boot time
done. HV should save itself a copy if the content is still needed after
init_paging().
Tracked-On: #2349
Signed-off-by: Chaohong guo <chaohong.guo@intel.com>
Reviewed-by: Zheng, Gen <gen.zheng@intel.com>
Current EFI boot related structs are allocated right after HV binary, and
marked as EfiResered. In fact, the memory should be reclaimed and
managed by HV once EFI boot is done, Moreover, HV is doing 2M alignment,
which is overwriting this allocated memory. This patch is to allocate EFI
info from EFI allocation pool and marked as EFiloaderdata, so that the
memory is used by HV, also, it will reduce the fragmentation.
Tracked-On: #2349
Signed-off-by: Chaohong guo <chaohong.guo@intel.com>
Reviewed-by: Eddie Dong <eddie.dong@intel.com>
Reviewed-by: Zheng, Gen <gen.zheng@intel.com>
Hypervisor does memory alignment during initializing paging(init_paging)
although the starting address allocated by EFI stub doesn't meet the
requirement. As a result, HV might overwrite some memory belong to UEFI
BIOS. The patch introduces a new routine to fix the issue.
Tracked-On: #2349
Signed-off-by: Chaohong guo <chaohong.guo@intel.com>
Reviewed-by: Zheng, Gen <gen.zheng@intel.com>
Reviewed-by: Zide Chen <zide.chen@intel.com>
The build environment might want to pass extra CFLAGS or LDFLAGS to the build of
the acrn tools. With conventional build systems like automake, Meson, etc this
is possible by just setting CFLAGS and LDFLAGS in the environment.
However as the tools are built using bare Makefiles, these environment variables
are overwritten. Respect them by renaming the variables in the Makefiles to
e.g. LOG_CFLAGS and adding CFLAGS to that.
Tracked-On: #2316
Signed-off-by: Ross Burton <ross.burton@intel.com>
-- fix "Macro parameter not in brackets" for PLUG_CPU(n)
-- fix "Array has no bounds specified" for vm_configs[]
Tracked-On: #861
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Using 'make' directly means that the jobserver environment variables don't get
passed down, so sub-builds for example don't use -j.
This is documented as the wrong thing to do:
https://www.gnu.org/software/make/manual/html_node/MAKE-Variable.html#MAKE-Variable
Use $(MAKE) instead, and compile times drastically improve:
acrn-devicemodel do_compile -13.5s -89.6% 15.0s -> 1.6s
Tracked-On: #2370
Signed-off-by: Ross Burton <ross.burton@intel.com>
This is the following patch after removing pending_pre_work, it make sure
all io emulation is done on its own cpu.
For hv emulated request, it follows the steps:
hv_emulate_pio->emulate_pio_complete
hv_emulate_mmio->emulate_mmio_complete
For dm emulated request, it follows the steps:
acrn_insert_request->dm_emulate_io_complete
while in acrn_insert_request, it could trigger scheduling out then resume by
hcall_notify_ioreq_finish, or busy wait for ioreq completion if polling mode
is enabled.
Tracked-On: #2394
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Xu, Anthony <anthony.xu@intel.com>
Now we do not need pending_pre_work anymore, as we can make sure IO request
VCPU resume from where it paused.
Now only three fixed points will try to do schedule:
- vcpu_thread: before vm entry, will check reschedule flag and to it if needed
- default_idle: loop check reschedule flag to see if need switch out
- io request: if IO REQ need DM's handle, it will schedule out
Tracked-On: #2394
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Xu, Anthony <anthony.xu@intel.com>
this patch added full context switch support for scheduling, it make sure
each VCPU has its own stack, and added host_sp field in struct sched_object
to record host stack pointer for each switch out object.
Arch related function arch_switch_to is added for context switch.
To benefit debugging, a name[] field is also added into struct sched_object.
Tracked-On: #2394
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Xu, Anthony <anthony.xu@intel.com>
To support full context switch scheduling, each physical
CPU must have its own idle sched_obj.
Tracked-On: #2394
Signed-off-by: Jason Chen CJ <jason.cj.chen@intel.com>
Acked-by: Xu, Anthony <anthony.xu@intel.com>
The existing .config may have missed or conflicting values if the file is
manually edited or Kconfig is changed. Kconfiglib will assign default values to
missed symbols and resolve conflicts when loading .config, but the current
silentoldconfig.py fails to write the updated symbol values back to .config.
This patch checks if any missed or conflicting symbol value is fixed by
Kconfiglib and write the updated values back to .config if necessary.
Tracked-On: #2371
Signed-off-by: Junjie Mao <junjie.mao@intel.com>
For virtio-blk, when the backend is a regular file, the discard
and
is implemented by fallocate(), but this function will not wait for
the discard command handled by disk.
So, add fdatasync to make sure the DISCARD is executed
synchronously.
Tracked-On: #2395
Signed-off-by: Conghui Chen <conghui.chen@intel.com>
Acked-by: Wang Yu <yu1.wang@intel.com>
vHPET is used as a source of system timer by UEFI (e.g. OVMF).
This provides an alternative to using vPIT, which OVMF assumes is always
connected to vPIC.
This is ported from Bhyve, with a few changes:
- move to user space, using acrn_timer
- enable timers only when necessary
Origin: FreeBSD
License: BSD-3-Clause
URL: https://svnweb.freebsd.org/
commit: 326257
Purpose: Adding vHPET support.
Maintained-by: External
Tracked-On: #2319
Signed-off-by: Peter Fang <peter.fang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
It is possible for multiple timeouts to occur in one mevent epoll
iteration. Providing the number of timer expirations to the timer
callback handlers can be useful. E.g., this could improve emulation of
timing-sensitive hardware components.
Tracked-On: #2319
Signed-off-by: Peter Fang <peter.fang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Allow specifying the initial expiration in absolute time.
Tracked-On: #2319
Signed-off-by: Peter Fang <peter.fang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
It was designed to be used in non-blocking mode to prevent the mevent
thread from blocking itself indefinitely, but it was created in
blocking mode.
Tracked-On: #2319
Signed-off-by: Peter Fang <peter.fang@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
This patch adds the dependency of $(LIB_FLAGS) when building .out file.
So that the target is updated whenever there is a update in
debug/release library.
Tracked-On: #2389
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
Reviewed-by: Binbin Wu <binbin.wu@intel.com>
For binutils <= 2.26:
'U' suffix for unsigned constant is not accepted by binutils assembler.
This patch separates the MACRO definition for ASSEMBLER
and non-ASSEMBLER code in idt.h
Tracked-On: #2200
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
separate the private macro/structure/functions, which are just used
in instr_emul.c;
and move instr_emul.h to public include path.
Tracked-On: #1842
Signed-off-by: Minggui Cao <minggui.cao@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
