From mboxrd@z Thu Jan  1 00:00:00 1970
From: Julien Grall <julien.grall@linaro.org>
Subject: Re: [PATCH] xen: arm: arm64: Adding VFP save/restore
 support.
Date: Thu, 06 Feb 2014 12:44:04 +0000
Message-ID: <52F38394.50405@linaro.org>
References: <1391671722-16127-1-git-send-email-pranavkumar@linaro.org>
Mime-Version: 1.0
Content-Type: text/plain; charset="us-ascii"; Format="flowed"
Content-Transfer-Encoding: 7bit
Return-path: <xen-devel-bounces@lists.xen.org>
In-Reply-To: <1391671722-16127-1-git-send-email-pranavkumar@linaro.org>
List-Unsubscribe: <http://lists.xen.org/cgi-bin/mailman/options/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=unsubscribe>
List-Post: <mailto:xen-devel@lists.xen.org>
List-Help: <mailto:xen-devel-request@lists.xen.org?subject=help>
List-Subscribe: <http://lists.xen.org/cgi-bin/mailman/listinfo/xen-devel>,
	<mailto:xen-devel-request@lists.xen.org?subject=subscribe>
Sender: xen-devel-bounces@lists.xen.org
Errors-To: xen-devel-bounces@lists.xen.org
To: Pranavkumar Sawargaonkar <pranavkumar@linaro.org>, xen-devel@lists.xen.org
Cc: stefano.stabellini@citrix.com, patches@apm.com, ian.campbell@citrix.com, Anup Patel <anup.patel@linaro.org>, patches@linaro.org
List-Id: xen-devel@lists.xenproject.org

Hello,

On 06/02/14 07:28, Pranavkumar Sawargaonkar wrote:
> This patch adds VFP save/restore support form arm64 across context switch.
>
> Signed-off-by: Pranavkumar Sawargaonkar <pranavkumar@linaro.org>
> Signed-off-by: Anup Patel <anup.patel@linaro.org>
> ---
>   xen/arch/arm/arm64/vfp.c        |   49 +++++++++++++++++++++++++++++++++++++++
>   xen/include/asm-arm/arm64/vfp.h |    4 ++++
>   2 files changed, 53 insertions(+)
>
> diff --git a/xen/arch/arm/arm64/vfp.c b/xen/arch/arm/arm64/vfp.c
> index 74e6a50..8c1479a 100644
> --- a/xen/arch/arm/arm64/vfp.c
> +++ b/xen/arch/arm/arm64/vfp.c
> @@ -1,13 +1,62 @@
>   #include <xen/sched.h>
>   #include <asm/processor.h>
> +#include <asm/cpufeature.h>
>   #include <asm/vfp.h>
>
>   void vfp_save_state(struct vcpu *v)
>   {
>       /* TODO: implement it */
> +    if ( !cpu_has_fp )
> +        return;
> +
> +    asm volatile("stp q0, q1, [%0, #16 * 0]\n\t"
> +                 "stp q2, q3, [%0, #16 * 2]\n\t"
> +                 "stp q4, q5, [%0, #16 * 4]\n\t"
> +                 "stp q6, q7, [%0, #16 * 6]\n\t"
> +                 "stp q8, q9, [%0, #16 * 8]\n\t"
> +                 "stp q10, q11, [%0, #16 * 10]\n\t"
> +                 "stp q12, q13, [%0, #16 * 12]\n\t"
> +                 "stp q14, q15, [%0, #16 * 14]\n\t"
> +                 "stp q16, q17, [%0, #16 * 16]\n\t"
> +                 "stp q18, q19, [%0, #16 * 18]\n\t"
> +                 "stp q20, q21, [%0, #16 * 20]\n\t"
> +                 "stp q22, q23, [%0, #16 * 22]\n\t"
> +                 "stp q24, q25, [%0, #16 * 24]\n\t"
> +                 "stp q26, q27, [%0, #16 * 26]\n\t"
> +                 "stp q28, q29, [%0, #16 * 28]\n\t"
> +                 "stp q30, q31, [%0, #16 * 30]\n\t"
> +                 :: "r" ((char *)(&v->arch.vfp.fpregs)): "memory");

I remember we had a discussion when I have implemented vfp context 
switch for arm32 for the memory constraints 
(http://lists.xen.org/archives/html/xen-devel/2013-06/msg00110.html).

I think you should use "=Q" also here to avoid cloberring the whole memory.

At the same time, is it necessary the (char *)?

> +
> +    v->arch.vfp.fpsr = READ_SYSREG32(FPSR);
> +    v->arch.vfp.fpcr = READ_SYSREG32(FPCR);
> +    v->arch.vfp.fpexc32_el2 = READ_SYSREG32(FPEXC32_EL2);
>   }
>
>   void vfp_restore_state(struct vcpu *v)
>   {
>       /* TODO: implement it */
> +    if ( !cpu_has_fp )
> +        return;
> +
> +    asm volatile("ldp q0, q1, [%0, #16 * 0]\n\t"
> +                 "ldp q2, q3, [%0, #16 * 2]\n\t"
> +                 "ldp q4, q5, [%0, #16 * 4]\n\t"
> +                 "ldp q6, q7, [%0, #16 * 6]\n\t"
> +                 "ldp q8, q9, [%0, #16 * 8]\n\t"
> +                 "ldp q10, q11, [%0, #16 * 10]\n\t"
> +                 "ldp q12, q13, [%0, #16 * 12]\n\t"
> +                 "ldp q14, q15, [%0, #16 * 14]\n\t"
> +                 "ldp q16, q17, [%0, #16 * 16]\n\t"
> +                 "ldp q18, q19, [%0, #16 * 18]\n\t"
> +                 "ldp q20, q21, [%0, #16 * 20]\n\t"
> +                 "ldp q22, q23, [%0, #16 * 22]\n\t"
> +                 "ldp q24, q25, [%0, #16 * 24]\n\t"
> +                 "ldp q26, q27, [%0, #16 * 26]\n\t"
> +                 "ldp q28, q29, [%0, #16 * 28]\n\t"
> +                 "ldp q30, q31, [%0, #16 * 30]\n\t"
> +                 :: "r" ((char *)(&v->arch.vfp.fpregs)): "memory");

Same here.

Regards,

-- 
Julien Grall