linux-arm-kernel.lists.infradead.org archive mirror
 help / color / mirror / Atom feed
From: cov@codeaurora.org (Christopher Covington)
To: linux-arm-kernel@lists.infradead.org
Subject: [PATCH 24/29] arm64: KVM: 32bit conditional execution emulation
Date: Mon, 18 Mar 2013 13:04:01 -0400	[thread overview]
Message-ID: <51474901.1010404@codeaurora.org> (raw)
In-Reply-To: <1362455265-24165-25-git-send-email-marc.zyngier@arm.com>

Hi Marc,

On 03/04/2013 10:47 PM, Marc Zyngier wrote:
> As conditionnal instructions can trap on AArch32, add the thinest

conditional

> possible emulation layer to keep 32bit guests happy.

[...]

> diff --git a/arch/arm64/kvm/emulate.c b/arch/arm64/kvm/emulate.c
> new file mode 100644
> index 0000000..6b3dbc3
> --- /dev/null
> +++ b/arch/arm64/kvm/emulate.c
> @@ -0,0 +1,154 @@
> +/*
> + * (not much of an) Emulation layer for 32bit guests.
> + *
> + * Copyright (C) 2012 - Virtual Open Systems and Columbia University
> + * Author: Christoffer Dall <c.dall@virtualopensystems.com>
> + *
> + * This program is free software: you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program.  If not, see <http://www.gnu.org/licenses/>.
> + */
> +
> +#include <linux/kvm_host.h>
> +#include <asm/kvm_emulate.h>
> +
> +/*
> + * stolen from arch/arm/kernel/opcodes.c
> + *
> + * condition code lookup table
> + * index into the table is test code: EQ, NE, ... LT, GT, AL, NV
> + *
> + * bit position in short is condition code: NZCV
> + */
> +static const unsigned short cc_map[16] = {
> +	0xF0F0,			/* EQ == Z set            */
> +	0x0F0F,			/* NE                     */
> +	0xCCCC,			/* CS == C set            */
> +	0x3333,			/* CC                     */
> +	0xFF00,			/* MI == N set            */
> +	0x00FF,			/* PL                     */
> +	0xAAAA,			/* VS == V set            */
> +	0x5555,			/* VC                     */
> +	0x0C0C,			/* HI == C set && Z clear */
> +	0xF3F3,			/* LS == C clear || Z set */
> +	0xAA55,			/* GE == (N==V)           */
> +	0x55AA,			/* LT == (N!=V)           */
> +	0x0A05,			/* GT == (!Z && (N==V))   */
> +	0xF5FA,			/* LE == (Z || (N!=V))    */
> +	0xFFFF,			/* AL always              */
> +	0			/* NV                     */
> +};
> +
> +static int kvm_vcpu_get_condition(struct kvm_vcpu *vcpu)
> +{
> +	u32 esr = kvm_vcpu_get_hsr(vcpu);
> +
> +	if (esr & ESR_EL2_CV)
> +		return (esr & ESR_EL2_COND) >> ESR_EL2_COND_SHIFT;
> +
> +	return -1;
> +}
> +
> +/*
> + * Check if a trapped instruction should have been executed or not.
> + */
> +bool kvm_condition_valid32(struct kvm_vcpu *vcpu)
> +{
> +	unsigned long cpsr;
> +	u32 cpsr_cond;
> +	int cond;
> +
> +	/* Top two bits non-zero?  Unconditional. */
> +	if (kvm_vcpu_get_hsr(vcpu) >> 30)
> +		return true;
> +
> +	/* Is condition field valid? */
> +	cond = kvm_vcpu_get_condition(vcpu);
> +	if (cond == 0xE)
> +		return true;
> +
> +	cpsr = *vcpu_cpsr(vcpu);
> +
> +	if (cond < 0) {
> +		/* This can happen in Thumb mode: examine IT state. */
> +		unsigned long it;
> +
> +		it = ((cpsr >> 8) & 0xFC) | ((cpsr >> 25) & 0x3);
> +
> +		/* it == 0 => unconditional. */
> +		if (it == 0)
> +			return true;
> +
> +		/* The cond for this insn works out as the top 4 bits. */
> +		cond = (it >> 4);
> +	}
> +
> +	cpsr_cond = cpsr >> 28;
> +
> +	if (!((cc_map[cond] >> cpsr_cond) & 1))
> +		return false;
> +
> +	return true;
> +}
> +
> +/**
> + * adjust_itstate - adjust ITSTATE when emulating instructions in IT-block
> + * @vcpu:	The VCPU pointer
> + *
> + * When exceptions occur while instructions are executed in Thumb IF-THEN
> + * blocks, the ITSTATE field of the CPSR is not advanved (updated), so we have
> + * to do this little bit of work manually. The fields map like this:
> + *
> + * IT[7:0] -> CPSR[26:25],CPSR[15:10]
> + */
> +static void kvm_adjust_itstate(struct kvm_vcpu *vcpu)
> +{
> +	unsigned long itbits, cond;
> +	unsigned long cpsr = *vcpu_cpsr(vcpu);
> +	bool is_arm = !(cpsr & COMPAT_PSR_T_BIT);
> +
> +	BUG_ON(is_arm && (cpsr & COMPAT_PSR_IT_MASK));
> +
> +	if (!(cpsr & COMPAT_PSR_IT_MASK))
> +		return;
> +
> +	cond = (cpsr & 0xe000) >> 13;
> +	itbits = (cpsr & 0x1c00) >> (10 - 2);
> +	itbits |= (cpsr & (0x3 << 25)) >> 25;
> +
> +	/* Perform ITAdvance (see page A-52 in ARM DDI 0406C) */
> +	if ((itbits & 0x7) == 0)
> +		itbits = cond = 0;
> +	else
> +		itbits = (itbits << 1) & 0x1f;
> +
> +	cpsr &= ~COMPAT_PSR_IT_MASK;
> +	cpsr |= cond << 13;
> +	cpsr |= (itbits & 0x1c) << (10 - 2);
> +	cpsr |= (itbits & 0x3) << 25;
> +	*vcpu_cpsr(vcpu) = cpsr;
> +}

Maybe I'm spoiled by the breadth of the 64-bit definitions, but I wonder if
fewer magic numbers in the 32-bit emulation might make it an easier read.

Christopher

-- 
Employee of Qualcomm Innovation Center, Inc.
Qualcomm Innovation Center, Inc. is a member of Code Aurora Forum, hosted by
the Linux Foundation

  reply	other threads:[~2013-03-18 17:04 UTC|newest]

Thread overview: 64+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-03-05  3:47 [PATCH 00/29] Port of KVM to arm64 Marc Zyngier
2013-03-05  3:47 ` [PATCH 01/29] arm64: KVM: define HYP and Stage-2 translation page flags Marc Zyngier
2013-03-05  3:47 ` [PATCH 02/29] arm64: KVM: HYP mode idmap support Marc Zyngier
2013-03-05  3:47 ` [PATCH 03/29] arm64: KVM: EL2 register definitions Marc Zyngier
2013-03-05  3:47 ` [PATCH 04/29] arm64: KVM: system register definitions for 64bit guests Marc Zyngier
2013-03-07 10:33   ` [kvmarm] " Alexander Graf
2013-03-08  3:23     ` Marc Zyngier
2013-03-12 13:20   ` Christopher Covington
2013-03-12 13:41     ` Christopher Covington
2013-03-12 13:50     ` Marc Zyngier
2013-03-05  3:47 ` [PATCH 05/29] arm64: KVM: Basic ESR_EL2 helpers and vcpu register access Marc Zyngier
2013-03-16  0:55   ` Geoff Levand
2013-03-05  3:47 ` [PATCH 06/29] arm64: KVM: fault injection into a guest Marc Zyngier
2013-03-12 13:20   ` Christopher Covington
2013-03-12 14:25     ` Marc Zyngier
2013-03-16  1:03   ` Geoff Levand
2013-03-05  3:47 ` [PATCH 07/29] arm64: KVM: architecture specific MMU backend Marc Zyngier
2013-03-05  3:47 ` [PATCH 08/29] arm64: KVM: user space interface Marc Zyngier
2013-03-07  8:09   ` Michael S. Tsirkin
2013-03-08  3:46     ` [kvmarm] " Marc Zyngier
2013-03-10  9:23       ` Michael S. Tsirkin
2013-03-05  3:47 ` [PATCH 09/29] arm64: KVM: system register handling Marc Zyngier
2013-03-07 10:30   ` [kvmarm] " Alexander Graf
2013-03-08  3:29     ` Marc Zyngier
2013-03-25  8:19     ` Marc Zyngier
2013-04-23 23:07       ` Christoffer Dall
2013-03-05  3:47 ` [PATCH 10/29] arm64: KVM: Cortex-A57 specific system registers handling Marc Zyngier
2013-03-13 18:30   ` Christopher Covington
2013-03-14 10:26     ` Marc Zyngier
2013-03-05  3:47 ` [PATCH 11/29] arm64: KVM: virtual CPU reset Marc Zyngier
2013-03-05  3:47 ` [PATCH 12/29] arm64: KVM: kvm_arch and kvm_vcpu_arch definitions Marc Zyngier
2013-03-12 17:30   ` Christopher Covington
2013-03-05  3:47 ` [PATCH 13/29] arm64: KVM: MMIO access backend Marc Zyngier
2013-03-05  3:47 ` [PATCH 14/29] arm64: KVM: guest one-reg interface Marc Zyngier
2013-03-12 17:31   ` Christopher Covington
2013-03-12 18:05     ` Marc Zyngier
2013-03-12 22:07       ` Christopher Covington
2013-03-13  7:48         ` Marc Zyngier
2013-03-13 20:34           ` Christopher Covington
2013-03-14  8:57             ` [kvmarm] " Peter Maydell
2013-03-20 20:06               ` Christopher Covington
2013-03-05  3:47 ` [PATCH 15/29] arm64: KVM: hypervisor initialization code Marc Zyngier
2013-03-05  3:47 ` [PATCH 16/29] arm64: KVM: HYP mode world switch implementation Marc Zyngier
2013-03-13 19:59   ` Christopher Covington
2013-03-20 20:04     ` Christopher Covington
2013-03-21 11:54       ` Marc Zyngier
2013-03-05  3:47 ` [PATCH 17/29] arm64: KVM: Exit handling Marc Zyngier
2013-03-05  3:47 ` [PATCH 18/29] arm64: KVM: Plug the VGIC Marc Zyngier
2013-03-05  3:47 ` [PATCH 19/29] arm64: KVM: Plug the arch timer Marc Zyngier
2013-03-05  3:47 ` [PATCH 20/29] arm64: KVM: PSCI implementation Marc Zyngier
2013-03-05  3:47 ` [PATCH 21/29] arm64: KVM: Build system integration Marc Zyngier
2013-03-05  3:47 ` [PATCH 22/29] arm64: KVM: define 32bit specific registers Marc Zyngier
2013-03-18 17:03   ` Christopher Covington
2013-03-05  3:47 ` [PATCH 23/29] arm64: KVM: 32bit GP register access Marc Zyngier
2013-03-16  0:24   ` Geoff Levand
2013-03-05  3:47 ` [PATCH 24/29] arm64: KVM: 32bit conditional execution emulation Marc Zyngier
2013-03-18 17:04   ` Christopher Covington [this message]
2013-03-05  3:47 ` [PATCH 25/29] arm64: KVM: 32bit handling of coprocessor traps Marc Zyngier
2013-03-05  3:47 ` [PATCH 26/29] arm64: KVM: 32bit coprocessor access for Cortex-A57 Marc Zyngier
2013-03-05  3:47 ` [PATCH 27/29] arm64: KVM: 32bit specific register world switch Marc Zyngier
2013-03-05  3:47 ` [PATCH 28/29] arm64: KVM: 32bit guest fault injection Marc Zyngier
2013-03-18 18:45   ` Christopher Covington
2013-03-05  3:47 ` [PATCH 29/29] arm64: KVM: enable initialization of a 32bit vcpu Marc Zyngier
2013-03-18 18:56   ` Christopher Covington

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=51474901.1010404@codeaurora.org \
    --to=cov@codeaurora.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).