linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: cdall@linaro.org (Christoffer Dall)
To: linux-arm-kernel@lists.infradead.org
Subject: [RFC 09/55] KVM: arm64: Set shadow EL1 registers for virtual EL2 execution
Date: Wed, 22 Feb 2017 12:19:52 +0100	[thread overview]
Message-ID: <20170222111952.GE26976@cbox> (raw)
In-Reply-To: <1483943091-1364-10-git-send-email-jintack@cs.columbia.edu>

On Mon, Jan 09, 2017 at 01:24:05AM -0500, Jintack Lim wrote:
> From: Christoffer Dall <christoffer.dall@linaro.org>
> 
> When entering virtual EL2, we need to reflect virtual EL2 register
> states to corresponding shadow EL1 registers. We can simply copy them if
> their formats are identical.  Otherwise, we need to convert EL2 register
> state to EL1 register state.
> 
> Signed-off-by: Christoffer Dall <christoffer.dall@linaro.org>
> Signed-off-by: Jintack Lim <jintack@cs.columbia.edu>
> ---
>  arch/arm64/kvm/context.c | 71 ++++++++++++++++++++++++++++++++++++++++++++++++

Looking at this again, I'm not sure 'context.c' is a very meaningful
name.

>  1 file changed, 71 insertions(+)
> 
> diff --git a/arch/arm64/kvm/context.c b/arch/arm64/kvm/context.c
> index acb4b1e..2e9e386 100644
> --- a/arch/arm64/kvm/context.c
> +++ b/arch/arm64/kvm/context.c
> @@ -17,6 +17,76 @@
>  
>  #include <linux/kvm_host.h>
>  #include <asm/kvm_emulate.h>
> +#include <asm/esr.h>
> +
> +struct el1_el2_map {
> +	enum vcpu_sysreg	el1;
> +	enum el2_regs		el2;
> +};
> +
> +/*
> + * List of EL2 registers which can be directly applied to EL1 registers to
> + * emulate running EL2 in EL1.  The EL1 registers here must either be trapped
> + * or paravirtualized in EL1.

This series doesn't deal with paravirtualizaion but only targets 8.3 so
we should clean up references to paravirtualization.

> + */
> +static const struct el1_el2_map el1_el2_map[] = {
> +	{ AMAIR_EL1, AMAIR_EL2 },
> +	{ MAIR_EL1, MAIR_EL2 },
> +	{ TTBR0_EL1, TTBR0_EL2 },
> +	{ ACTLR_EL1, ACTLR_EL2 },
> +	{ AFSR0_EL1, AFSR0_EL2 },
> +	{ AFSR1_EL1, AFSR1_EL2 },
> +	{ SCTLR_EL1, SCTLR_EL2 },
> +	{ VBAR_EL1, VBAR_EL2 },
> +};
> +
> +static inline u64 tcr_el2_ips_to_tcr_el1_ps(u64 tcr_el2)
> +{
> +	return ((tcr_el2 & TCR_EL2_PS_MASK) >> TCR_EL2_PS_SHIFT)
> +		<< TCR_IPS_SHIFT;
> +}
> +
> +static inline u64 cptr_el2_to_cpacr_el1(u64 cptr_el2)
> +{
> +	u64 cpacr_el1 = 0;
> +
> +	if (!(cptr_el2 & CPTR_EL2_TFP))
> +		cpacr_el1 |= CPACR_EL1_FPEN;
> +	if (cptr_el2 & CPTR_EL2_TTA)
> +		cpacr_el1 |= CPACR_EL1_TTA;
> +
> +	return cpacr_el1;
> +}
> +
> +static void create_shadow_el1_sysregs(struct kvm_vcpu *vcpu)
> +{
> +	u64 *s_sys_regs = vcpu->arch.ctxt.shadow_sys_regs;
> +	u64 *el2_regs = vcpu->arch.ctxt.el2_regs;
> +	u64 tcr_el2;
> +	int i;
> +
> +	for (i = 0; i < ARRAY_SIZE(el1_el2_map); i++) {
> +		const struct el1_el2_map *map = &el1_el2_map[i];
> +
> +		s_sys_regs[map->el1] = el2_regs[map->el2];
> +	}
> +
> +	tcr_el2 = el2_regs[TCR_EL2];
> +	s_sys_regs[TCR_EL1] =
> +		TCR_EPD1 |	/* disable TTBR1_EL1 */
> +		((tcr_el2 & TCR_EL2_TBI) ? TCR_TBI0 : 0) |
> +		tcr_el2_ips_to_tcr_el1_ps(tcr_el2) |
> +		(tcr_el2 & TCR_EL2_TG0_MASK) |
> +		(tcr_el2 & TCR_EL2_ORGN0_MASK) |
> +		(tcr_el2 & TCR_EL2_IRGN0_MASK) |
> +		(tcr_el2 & TCR_EL2_T0SZ_MASK);
> +
> +	/* Rely on separate VMID for VA context, always use ASID 0 */
> +	s_sys_regs[TTBR0_EL1] &= ~GENMASK_ULL(63, 48);
> +	s_sys_regs[TTBR1_EL1] = 0;
> +
> +	s_sys_regs[CPACR_EL1] = cptr_el2_to_cpacr_el1(el2_regs[CPTR_EL2]);
> +}
>  
>  /**
>   * kvm_arm_setup_shadow_state -- prepare shadow state based on emulated mode
> @@ -37,6 +107,7 @@ void kvm_arm_setup_shadow_state(struct kvm_vcpu *vcpu)
>  		else
>  			ctxt->hw_pstate |= PSR_MODE_EL1t;
>  
> +		create_shadow_el1_sysregs(vcpu);
>  		ctxt->hw_sys_regs = ctxt->shadow_sys_regs;
>  		ctxt->hw_sp_el1 = ctxt->el2_regs[SP_EL2];
>  	} else {
> -- 
> 1.9.1
> 
> 

  reply	other threads:[~2017-02-22 11:19 UTC|newest]

Thread overview: 111+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2017-01-09  6:23 [RFC 00/55] Nested Virtualization on KVM/ARM Jintack Lim
2017-01-09  6:23 ` [RFC 01/55] arm64: Add missing TCR hw defines Jintack Lim
2017-01-09  6:23 ` [RFC 02/55] KVM: arm64: Add nesting config option Jintack Lim
2017-01-09  6:23 ` [RFC 03/55] KVM: arm64: Add KVM nesting feature Jintack Lim
2017-01-09  6:24 ` [RFC 04/55] KVM: arm64: Allow userspace to set PSR_MODE_EL2x Jintack Lim
2017-01-09  6:24 ` [RFC 05/55] KVM: arm64: Add vcpu_mode_el2 primitive to support nesting Jintack Lim
2017-01-09  6:24 ` [RFC 06/55] KVM: arm64: Add EL2 execution context for nesting Jintack Lim
2017-02-22 11:10   ` Christoffer Dall
2017-06-26 14:33     ` Jintack Lim
2017-07-03  9:03       ` Christoffer Dall
2017-07-03  9:32         ` Marc Zyngier
2017-07-03  9:54           ` Christoffer Dall
2017-07-03 14:44             ` Jintack Lim
2017-07-03 15:30               ` Christoffer Dall
2017-01-09  6:24 ` [RFC 07/55] KVM: arm/arm64: Add virtual EL2 state emulation framework Jintack Lim
2017-02-22 11:12   ` Christoffer Dall
2017-06-01 20:05   ` Bandan Das
2017-06-02 11:51     ` Christoffer Dall
2017-06-02 17:36       ` Bandan Das
2017-06-02 19:06         ` Christoffer Dall
2017-06-02 19:25           ` Bandan Das
2017-01-09  6:24 ` [RFC 08/55] KVM: arm64: Set virtual EL2 context depending on the guest exception level Jintack Lim
2017-02-22 11:14   ` Christoffer Dall
2017-06-01 20:22   ` Bandan Das
2017-06-02  8:48     ` Marc Zyngier
2017-01-09  6:24 ` [RFC 09/55] KVM: arm64: Set shadow EL1 registers for virtual EL2 execution Jintack Lim
2017-02-22 11:19   ` Christoffer Dall [this message]
2017-01-09  6:24 ` [RFC 10/55] KVM: arm64: Synchronize EL1 system registers on virtual EL2 entry and exit Jintack Lim
2017-06-06 20:16   ` Bandan Das
2017-06-07  4:26     ` Jintack Lim
2017-01-09  6:24 ` [RFC 11/55] KVM: arm64: Emulate taking an exception to the guest hypervisor Jintack Lim
2017-02-22 11:28   ` Christoffer Dall
2017-06-06 20:21   ` Bandan Das
2017-06-06 20:38     ` Jintack Lim
2017-06-06 22:07       ` Bandan Das
2017-06-06 23:16         ` Jintack Lim
2017-06-07 17:21           ` Bandan Das
2017-01-09  6:24 ` [RFC 12/55] KVM: arm64: Handle EL2 register access traps Jintack Lim
2017-02-22 11:30   ` Christoffer Dall
2017-02-22 11:31   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 13/55] KVM: arm64: Handle eret instruction traps Jintack Lim
2017-01-09  6:24 ` [RFC 14/55] KVM: arm64: Take account of system " Jintack Lim
2017-02-22 11:34   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 15/55] KVM: arm64: Trap EL1 VM register accesses in virtual EL2 Jintack Lim
2017-01-09  6:24 ` [RFC 16/55] KVM: arm64: Forward VM reg traps to the guest hypervisor Jintack Lim
2017-02-22 11:39   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 17/55] KVM: arm64: Trap SPSR_EL1, ELR_EL1 and VBAR_EL1 in virtual EL2 Jintack Lim
2017-02-22 11:40   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 18/55] KVM: arm64: Forward traps due to HCR_EL2.NV1 bit to the guest hypervisor Jintack Lim
2017-02-22 11:41   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 19/55] KVM: arm64: Trap CPACR_EL1 access in virtual EL2 Jintack Lim
2017-01-09  6:24 ` [RFC 20/55] KVM: arm64: Forward CPACR_EL1 traps to the guest hypervisor Jintack Lim
2017-01-09  6:24 ` [RFC 21/55] KVM: arm64: Forward HVC instruction " Jintack Lim
2017-02-22 11:47   ` Christoffer Dall
2017-06-26 15:21     ` Jintack Lim
2017-07-03  9:08       ` Christoffer Dall
2017-07-03  9:31         ` Andrew Jones
2017-07-03  9:51           ` Christoffer Dall
2017-07-03 12:03             ` Will Deacon
2017-07-03 12:35               ` Marc Zyngier
2017-07-03 13:29         ` Jintack Lim
2017-01-09  6:24 ` [RFC 22/55] KVM: arm64: Handle PSCI call from the guest Jintack Lim
2017-01-09  6:24 ` [RFC 23/55] KVM: arm64: Forward WFX to the guest hypervisor Jintack Lim
2017-01-09  6:24 ` [RFC 24/55] KVM: arm64: Forward FP exceptions " Jintack Lim
2017-01-09  6:24 ` [RFC 25/55] KVM: arm/arm64: Let vcpu thread modify its own active state Jintack Lim
2017-02-22 12:27   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 26/55] KVM: arm/arm64: Add VGIC data structures for the nesting Jintack Lim
2017-01-09  6:24 ` [RFC 27/55] KVM: arm/arm64: Emulate GICH interface on GICv2 Jintack Lim
2017-02-22 13:06   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 28/55] KVM: arm/arm64: Prepare vgic state for the nested VM Jintack Lim
2017-02-22 13:12   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 29/55] KVM: arm/arm64: Set up the prepared vgic state Jintack Lim
2017-01-09  6:24 ` [RFC 30/55] KVM: arm/arm64: Inject irqs to the guest hypervisor Jintack Lim
2017-02-22 13:16   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 31/55] KVM: arm/arm64: Inject maintenance interrupts " Jintack Lim
2017-02-22 13:19   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 32/55] KVM: arm/arm64: register GICH iodev for " Jintack Lim
2017-02-22 13:21   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 33/55] KVM: arm/arm64: Remove unused params in mmu functions Jintack Lim
2017-01-09  6:24 ` [RFC 34/55] KVM: arm/arm64: Abstract stage-2 MMU state into a separate structure Jintack Lim
2017-01-09  6:24 ` [RFC 35/55] KVM: arm/arm64: Support mmu for the virtual EL2 execution Jintack Lim
2017-02-22 13:38   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 36/55] KVM: arm64: Invalidate virtual EL2 TLB entries when needed Jintack Lim
2017-01-09  6:24 ` [RFC 37/55] KVM: arm64: Setup vttbr_el2 on each VM entry Jintack Lim
2017-01-09  6:24 ` [RFC 38/55] KVM: arm/arm64: Make mmu functions non-static Jintack Lim
2017-01-09  6:24 ` [RFC 39/55] KVM: arm/arm64: Add mmu context for the nesting Jintack Lim
2017-02-22 13:34   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 40/55] KVM: arm/arm64: Handle vttbr_el2 write operation from the guest hypervisor Jintack Lim
2017-02-22 17:59   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 41/55] KVM: arm/arm64: Unmap/flush shadow stage 2 page tables Jintack Lim
2017-02-22 18:09   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 42/55] KVM: arm64: Implement nested Stage-2 page table walk logic Jintack Lim
2017-01-09  6:24 ` [RFC 43/55] KVM: arm/arm64: Handle shadow stage 2 page faults Jintack Lim
2017-01-09  6:24 ` [RFC 44/55] KVM: arm/arm64: Move kvm_is_write_fault to header file Jintack Lim
2017-01-09  6:24 ` [RFC 45/55] KVM: arm64: KVM: Inject stage-2 page faults Jintack Lim
2017-01-09  6:24 ` [RFC 46/55] KVM: arm64: Add more info to the S2 translation result Jintack Lim
2017-01-09  6:24 ` [RFC 47/55] KVM: arm/arm64: Forward the guest hypervisor's stage 2 permission faults Jintack Lim
2017-02-22 18:15   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 48/55] KVM: arm64: Emulate TLBI instruction Jintack Lim
2017-01-09  6:24 ` [RFC 49/55] KVM: arm64: Fixes to toggle_cache for nesting Jintack Lim
2017-01-09  6:24 ` [RFC 50/55] KVM: arm/arm64: Abstract kvm_phys_addr_ioremap() function Jintack Lim
2017-01-09  6:24 ` [RFC 51/55] KVM: arm64: Expose physical address of vcpu interface Jintack Lim
2017-01-09  6:24 ` [RFC 52/55] KVM: arm/arm64: Create a vcpu mapping for the nested VM Jintack Lim
2017-01-09  6:24 ` [RFC 53/55] KVM: arm64: Reflect shadow VMPIDR_EL2 value to MPIDR_EL1 Jintack Lim
2017-01-09  6:24 ` [RFC 54/55] KVM: arm/arm64: Adjust virtual offset considering nesting Jintack Lim
2017-02-22 19:28   ` Christoffer Dall
2017-01-09  6:24 ` [RFC 55/55] KVM: arm64: Enable nested virtualization Jintack Lim
2017-01-09 15:05 ` [RFC 00/55] Nested Virtualization on KVM/ARM David Hildenbrand
2017-01-10 16:18   ` Jintack Lim
2017-02-22 18:23 ` Christoffer Dall
2017-02-24 10:28   ` Jintack Lim

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20170222111952.GE26976@cbox \
    --to=cdall@linaro.org \
    --cc=linux-arm-kernel@lists.infradead.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).