Commit Graph

15 Commits

Author SHA1 Message Date
Kaige Fu
a85d11ca7a HV: Add prefix 'p' before 'cpu' to physical cpu related functions
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>
2019-04-24 10:50:28 +08:00
li bing
25741b62db HV: fix the issue of ACRN_REQUEST_EXCP flag is not cleared.
the problem is : System will crash when run crashme.
The root cause of this problem is that when the ACRN_REQUEST_EXCP flag is set by calling
the vcpu_make_request function, the flag is not cleared.
Add the following statement to the vcpu_inject_exception function to fix the problem:
bitmap_test_and_clear_lock(ACRN_REQUEST_EXCP, &vcpu->arch.pending_req);
Tested that one night, there was no crash.

Tracked-On: #2527
Signed-off-by: bing.li<bingx.li@intel.com>
Acked-by:      Eddie Dong<eddie.dong@intel.com>
2019-04-23 15:17:13 +08:00
Li, Fei1
0c347e607a hv: vlapic: wrap APICv check pending delivery interrupt
When in fully APICv mode, we enable VID. All pending delivery interrupts
will inject to VM before VM entry. So there is no pending delivery interrupt.
However, if VID is not enabled, we can only inject pending delivery interrupt
one by one. So we always need to do this check.

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-04-23 15:16:56 +08:00
Li, Fei1
037fffc203 hv: vlapic: wrap APICv inject interrupt API
apicv_advanced_inject_intr is used if APICv fully features are supported,
it uses PIR to inject interrupt. otherwise, apicv_basic_inject_intr is used.
it will use VMCS INTR INFO field to inject irq.

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-04-23 15:16:56 +08:00
Li, Fei1
1db8123c2d hv: virq: refine pending event inject coding style
In order to combine vlapic_apicv_inject_pir and vcpu_inject_vlapic_int.

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-04-23 15:16:56 +08:00
Li, Fei1
fde2c07c0a hv: vlapic: minor fix about APICv inject interrupt
When VID is enabeld, we should always inejct the pending interrupts when vm enter.

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-04-23 15:16:56 +08:00
Li, Fei1
56acaacc29 hv: vlapic: add TPR below threshold implement
Add TPR below threshold implement for "Virtual-interrupt delivery" not support.
Windows will use it to delay interrupt handle.

Complete all the interrupts in IRR as long as they are higher priority than
current TPR. Once current IRR priority is less than current TPR enable TPR
threshold to IRR, so that if guest reduces the TPR threshold, it would be good
to take below TPR threshold exit and let interrupts to go thru.

Tracked-On: #1842
Signed-off-by: Zheng, Gen <gen.zheng@intel.com>
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-04-19 11:11:54 +08:00
Sainath Grandhi
f22347346e hv: Remove separate interrupt routine for pre-launched VMs
As vector re-mapping is enabled for pre-launched/partition mode VMs,
there is no more need for separate interrupt routine i.e.
partition_mode_dispatch_interrupt.

Tracked-On: #2879
Signed-off-by: Sainath Grandhi <sainath.grandhi@intel.com>
Reviewed-by: Eddie Dong <eddie.dong@intel.com>
2019-04-04 17:51:36 +08:00
Li, Fei1
f769f7457b hv: vlapic: add combined constraint for APICv
Add two functions to combine constraint for APICv:
is_apicv_basic_feature_supported: check the physical platform whether support
"Use TPR shadow", "Virtualize APIC accesses" and "Virtualize x2APIC mode"
is_apicv_advanced_feature_supported: check the physical platform whether support
"APIC-register virtualization", "Virtual-interrupt delivery" and
"Process posted interrupts".

If the physical platform only support APICv basic feature, enable "Use TPR shadow"
and "Virtualize APIC accesses" for xAPIC mode; enable "Use TPR shadow" and
"Virtualize x2APIC mode" for x2APIC. Otherwise, if the physical platform support
APICv advanced feature, enable APICv feature for xAPIC mode and x2APIC mode.

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-03-12 20:37:06 +08:00
Shiqing Gao
da14c96135 hv: simplify get_primary_vcpu and vcpu_from_vid
This patch simplifies `get_primary_vcpu` and `vcpu_from_vid`.
The target_vcpu could be get from the index directly.

Tracked-On: #1842
Signed-off-by: Shiqing Gao <shiqing.gao@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
2019-03-04 13:45:02 +08:00
Li, Fei1
2e60adef7c hv: vmcs: simplify update EOI-exit bitmap
1) The previous implementaion will recalculate the whole EOI-exit bitmap for
each RTE once the destination, trigger mode, delivery mode or vector of a RTE
has changed and update the EOI-exit bitmap for each vcpu of the VM.
In this patch, only set the corresponding bit of EOI-exit bitmap for
a vcpu when a level triggered interrupt has accepted in IRR or clear the
corresponding bit of EOI-exit bitmap for a vcpu when a dege triggered interrupt
has accepted in IRR which means only update a bit of EOI-exit bitmap in a vcpu
when updating TMR.
2) Rename set eoi_exit related API to set eoi_exit_bitmap.

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
2019-02-27 15:28:37 +08:00
Mingqiang Chi
b24a8a0f59 hv:cleanup header file for guest folder
cleanup arch/x86/guest, only include some necessary
header files, doesn't include hypervisor.h

Tracked-On: #1842
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
	modified:   arch/x86/guest/assign.c
	modified:   arch/x86/guest/ept.c
	modified:   arch/x86/guest/guest_memory.c
	modified:   arch/x86/guest/instr_emul.c
	modified:   arch/x86/guest/io_emul.c
	modified:   arch/x86/guest/pm.c
	modified:   arch/x86/guest/trusty.c
	modified:   arch/x86/guest/ucode.c
	modified:   arch/x86/guest/vcpu.c
	modified:   arch/x86/guest/vcpuid.c
	modified:   arch/x86/guest/virq.c
	modified:   arch/x86/guest/virtual_cr.c
	modified:   arch/x86/guest/vlapic.c
	modified:   arch/x86/guest/vm.c
	modified:   arch/x86/guest/vmcall.c
	modified:   arch/x86/guest/vmcs.c
	modified:   arch/x86/guest/vmexit.c
	modified:   arch/x86/guest/vmsr.c
	modified:   arch/x86/guest/vmtrr.c
	modified:   arch/x86/pm.c
	modified:   include/arch/x86/guest/assign.h
	modified:   include/arch/x86/guest/ept.h
	modified:   include/arch/x86/guest/guest_memory.h
	modified:   include/arch/x86/guest/instr_emul.h
	modified:   include/arch/x86/guest/io_emul.h
	modified:   include/arch/x86/guest/trusty.h
	modified:   include/arch/x86/guest/vcpu.h
	modified:   include/arch/x86/guest/vmcs.h
	modified:   include/arch/x86/io_req.h
	modified:   include/arch/x86/irq.h
	modified:   include/arch/x86/lapic.h
	modified:   include/arch/x86/mmu.h
	modified:   include/arch/x86/pgtable.h
	modified:   include/common/ptdev.h
	modified:   include/debug/console.h
2019-02-21 10:38:30 +08:00
Binbin Wu
827fffedda hv: exception: fault type exception should set resume flag in rflags
According to SDM 17.3.1.1, for any fault-class exception except a
debug exception generated in response to an instruction breakpoint,
the value pushed for RF is 1.

This patch set Resume Flag for fault class exceptions.

Tracked-On: #2405
Signed-off-by: Binbin Wu <binbin.wu@intel.com>
Acked-by: Anthony Xu <anthony.xu@intel.com>
Acked-by: Eddie Dong <eddie.dong@intel.com>
2019-02-20 13:21:45 +08:00
Li, Fei1
ce19dd423e hv: vlapic: make vlapic deliver interrupt related functions more readable
Rename vlapic_deliver_intr to vlapic_receive_intr: ioapic/msi device
deliver an interrupt to lapic.
Rename vlapic_pending_intr to vlapic_find_deliverable_intr: find a
deliverable interrupt which pending in irr and its priority large than ppr.
Rename vlapic_intr_accepted to vlapic_get_deliverable_intr: get the deliverable
interrupt from irr and set it in isr (which also raise ppr update)

Tracked-On: #1842
Signed-off-by: Li, Fei1 <fei1.li@intel.com>
2019-02-02 09:31:49 +08:00
Mingqiang Chi
2b2dbe43d7 hv:move some files to guest folder
move the following files to guest folder:
renamed:    arch/x86/assign.c -> arch/x86/guest/assign.c
renamed:    arch/x86/ept.c -> arch/x86/guest/ept.c
renamed:    arch/x86/io_emul.c -> arch/x86/guest/io_emul.c
renamed:    arch/x86/trusty.c -> arch/x86/guest/trusty.c
renamed:    arch/x86/virq.c -> arch/x86/guest/virq.c
renamed:    arch/x86/virtual_cr.c -> arch/x86/guest/virtual_cr.c
renamed:    arch/x86/vmcs.c -> arch/x86/guest/vmcs.c
renamed:    arch/x86/vmexit.c -> arch/x86/guest/vmexit.c
renamed:    arch/x86/vmx_asm.S -> arch/x86/guest/vmx_asm.S
renamed:    include/arch/x86/assign.h ->
	    include/arch/x86/guest/assign.h
renamed:    include/arch/x86/io_emul.h ->
	    include/arch/x86/guest/io_emul.h
renamed:    include/arch/x86/trusty.h ->
            include/arch/x86/guest/trusty.h
renamed:    include/arch/x86/virtual_cr.h ->
            include/arch/x86/guest/virtual_cr.h
renamed:    include/arch/x86/vmcs.h -> include/arch/x86/guest/vmcs.h
renamed:    include/arch/x86/vmexit.h ->
            include/arch/x86/guest/vmexit.h

After these files movement, all the files in arch/x86 are native
hardware related, and all the files in arch/x86/guest are
virtualiztion related.

Tracked-On: #1842
Signed-off-by: Mingqiang Chi <mingqiang.chi@intel.com>
Reviewed-by: Jason Chen CJ <jason.cj.chen@intel.com>
Reviewed-by: Eddie Dong <eddie.dong@intel.com>
2019-01-30 20:43:04 +08:00