linux-kernel - Re: [Xen-devel] [PATCH v2 09/16] arm/xen: Get event-channel irq through HVM

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Date:	Mon, 18 Jan 2016 14:38:27 +0000
From:	Stefano Stabellini <stefano.stabellini@...citrix.com>
To:	Shannon Zhao <zhaoshenglong@...wei.com>
CC:	<linux-arm-kernel@...ts.infradead.org>,
	<ard.biesheuvel@...aro.org>, <stefano.stabellini@...rix.com>,
	<david.vrabel@...rix.com>, <mark.rutland@....com>,
	<devicetree@...r.kernel.org>, <linux-efi@...r.kernel.org>,
	<catalin.marinas@....com>, <will.deacon@....com>,
	<linux-kernel@...r.kernel.org>, <xen-devel@...ts.xen.org>,
	<julien.grall@...rix.com>, <shannon.zhao@...aro.org>,
	<peter.huangpeng@...wei.com>
Subject: Re: [Xen-devel] [PATCH v2 09/16] arm/xen: Get event-channel irq
 through HVM_PARAM when booting with ACPI

On Fri, 15 Jan 2016, Shannon Zhao wrote:
> From: Shannon Zhao <shannon.zhao@...aro.org>
> 
> When booting with ACPI, it could get the event-channel irq through
> HVM_PARAM_CALLBACK_IRQ.
> 
> Signed-off-by: Shannon Zhao <shannon.zhao@...aro.org>
> ---
>  arch/arm/xen/enlighten.c | 30 +++++++++++++++++++++++++++++-
>  1 file changed, 29 insertions(+), 1 deletion(-)
> 
> diff --git a/arch/arm/xen/enlighten.c b/arch/arm/xen/enlighten.c
> index 6d90a62..4e78c49 100644
> --- a/arch/arm/xen/enlighten.c
> +++ b/arch/arm/xen/enlighten.c
> @@ -29,6 +29,7 @@
>  #include <linux/time64.h>
>  #include <linux/timekeeping.h>
>  #include <linux/timekeeper_internal.h>
> +#include <linux/acpi.h>
>  
>  #include <linux/mm.h>
>  
> @@ -270,7 +271,34 @@ static int __init xen_guest_init(void)
>  	if (!xen_domain())
>  		return 0;
>  
> -	xen_events_irq = irq_of_parse_and_map(xen_node, 0);
> +	if (!acpi_disabled) {
> +#ifdef CONFIG_ACPI

I think it would be best to move the acpi specific and DT specific
initializations out of xen_guest_init. Then at the beginning of
xen_guest_init, call only the one that applies.  The acpi init function
could be ifdef'ed CONFIG_ACPI. Making sure that acpi_disabled is 1 when
!CONFIG_ACPI, we could even avoid all ifdef's from xen_guest_init.

The rest is good, thanks.


> +		struct xen_hvm_param a;
> +		int interrupt, trigger, polarity;
> +
> +		a.domid = DOMID_SELF;
> +		a.index = HVM_PARAM_CALLBACK_IRQ;
> +		xen_events_irq = 0;
> +		if (!HYPERVISOR_hvm_op(HVMOP_get_param, &a)) {
> +			if ((a.value >> 56) == 3) {
> +				interrupt = a.value & 0xff;
> +				trigger = ((a.value >> 8) & 0x1)
> +					  ? ACPI_EDGE_SENSITIVE
> +					  : ACPI_LEVEL_SENSITIVE;
> +				polarity = ((a.value >> 8) & 0x2)
> +					  ? ACPI_ACTIVE_LOW
> +					  : ACPI_ACTIVE_HIGH;
> +				xen_events_irq = acpi_register_gsi(NULL,
> +								   interrupt,
> +								   trigger,
> +								   polarity);
> +			}
> +		}
> +
> +#endif
> +	} else {
> +		xen_events_irq = irq_of_parse_and_map(xen_node, 0);
> +	}
> +
>  	if (!xen_events_irq) {
>  		pr_err("Xen event channel interrupt not found\n");
>  		return -ENODEV;