xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: Julien Grall <julien.grall@linaro.org>
To: vijay.kilari@gmail.com
Cc: Ian.Campbell@citrix.com, stefano.stabellini@eu.citrix.com,
	Prasun.Kapoor@caviumnetworks.com,
	vijaya.kumar@caviumnetworks.com, xen-devel@lists.xen.org,
	stefano.stabellini@citrix.com
Subject: Re: [RFC PATCH v1 01/10] xen/arm: make secondary gic init as notifier call
Date: Thu, 20 Mar 2014 12:48:18 +0000	[thread overview]
Message-ID: <532AE392.1080107@linaro.org> (raw)
In-Reply-To: <1395238631-2024-2-git-send-email-vijay.kilari@gmail.com>

Hello Vijay,

Thanks for your patch.

On 03/19/2014 02:17 PM, vijay.kilari@gmail.com wrote:
> From: Vijaya Kumar K <Vijaya.Kumar@caviumnetworks.com>
> 
> make gic init for secondary cpus as notifier call
> instead calling directly from secondary boot for
> each cpu. This makes secondary gic init generic and runtime.
> 
> Signed-off-by: Vijaya Kumar K <Vijaya.Kumar@caviumnetworks.com>
> ---
>  xen/arch/arm/gic.c        |   35 ++++++++++++++++++++++++++---------
>  xen/arch/arm/smpboot.c    |    3 +--
>  xen/include/asm-arm/gic.h |    2 --
>  3 files changed, 27 insertions(+), 13 deletions(-)
> 
> diff --git a/xen/arch/arm/gic.c b/xen/arch/arm/gic.c
> index 91a2982..4be0897 100644
> --- a/xen/arch/arm/gic.c
> +++ b/xen/arch/arm/gic.c
> @@ -20,6 +20,7 @@
>  #include <xen/config.h>
>  #include <xen/lib.h>
>  #include <xen/init.h>
> +#include <xen/cpu.h>
>  #include <xen/mm.h>
>  #include <xen/irq.h>
>  #include <xen/sched.h>
> @@ -380,6 +381,30 @@ static void __cpuinit gic_hyp_disable(void)
>      GICH[GICH_HCR] = 0;
>  }
>  
> +/* Set up the per-CPU parts of the GIC for a secondary CPU */
> +static int  __cpuinit gic_init_secondary_cpu(struct notifier_block *nfb,
> +                                        unsigned long action, void *hcpu)
> +{
> +    if (action == CPU_STARTING)
> +    {
> +        spin_lock(&gic.lock);
> +        gic_cpu_init();
> +        gic_hyp_init();
> +        spin_unlock(&gic.lock);
> +    }
> +    return NOTIFY_DONE;
> +}
> +

This is not the correct way to create a notifier block.

You should have a good similar to (see an example in common/timer.c):

static cpu_callback(struct notifier_block* nfb,
                    unsigned long action, void *hcpu)
{
     unsigned int cpu = (unsigned long)hcpu;

     switch ( action )
     case CPU_STARTING:
        gic_init_secondary_cpu();
        break;
     default:
        break;
     return NOTIFY_DONE;
}

> +static struct notifier_block gic_cpu_nb = { 
> +    .notifier_call = gic_init_secondary_cpu,
> +    .priority = 100 
> +};

> +static void gic_smp_init(void)
> +{
> +   register_cpu_notifier(&gic_cpu_nb);
> +}
> +

You don't need to create a separate function to register the notifier.
You can directly call it in gic_init.

>  int gic_irq_xlate(const u32 *intspec, unsigned int intsize,
>                    unsigned int *out_hwirq,
>                    unsigned int *out_type)
> @@ -469,6 +494,7 @@ void __init gic_init(void)
>      spin_lock_init(&gic.lock);
>      spin_lock(&gic.lock);
>  
> +    gic_smp_init();
>      gic_dist_init();
>      gic_cpu_init();
>      gic_hyp_init();
> @@ -524,15 +550,6 @@ void smp_send_state_dump(unsigned int cpu)
>      send_SGI_one(cpu, GIC_SGI_DUMP_STATE);
>  }
>  
> -/* Set up the per-CPU parts of the GIC for a secondary CPU */
> -void __cpuinit gic_init_secondary_cpu(void)
> -{
> -    spin_lock(&gic.lock);
> -    gic_cpu_init();
> -    gic_hyp_init();
> -    spin_unlock(&gic.lock);
> -}
> -
>  /* Shut down the per-CPU GIC interface */
>  void gic_disable_cpu(void)
>  {
> diff --git a/xen/arch/arm/smpboot.c b/xen/arch/arm/smpboot.c
> index a829957..765efcf 100644
> --- a/xen/arch/arm/smpboot.c
> +++ b/xen/arch/arm/smpboot.c
> @@ -283,7 +283,7 @@ void __cpuinit start_secondary(unsigned long boot_phys_offset,
>  
>      mmu_init_secondary_cpu();
>  
> -    gic_init_secondary_cpu();
> +    notify_cpu_starting(cpuid);

Can you explain why it's safe to move notify_cpu_starting earlier?

>  
>      init_secondary_IRQ();
>  
> @@ -297,7 +297,6 @@ void __cpuinit start_secondary(unsigned long boot_phys_offset,
>      setup_cpu_sibling_map(cpuid);
>  
>      /* Run local notifiers */

Please move also the comment, it's part of notify_cpu_starting. It
doesn't make any sense alone here.

> -    notify_cpu_starting(cpuid);
>      wmb();

Regards,

-- 
Julien Grall

  reply	other threads:[~2014-03-20 12:48 UTC|newest]

Thread overview: 95+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2014-03-19 14:17 [RFC PATCH v1 00/10] xen/arm: Add GICv3 support vijay.kilari
2014-03-19 14:17 ` [RFC PATCH v1 01/10] xen/arm: make secondary gic init as notifier call vijay.kilari
2014-03-20 12:48   ` Julien Grall [this message]
2014-03-22  8:16     ` Vijay Kilari
2014-03-23 14:38       ` Julien Grall
2014-03-26 11:27         ` Vijay Kilari
2014-03-26 14:41           ` Julien Grall
2014-03-26 17:22             ` George Dunlap
2014-03-21 17:15   ` Ian Campbell
2014-03-22  8:32     ` Vijay Kilari
2014-03-22 13:54       ` Julien Grall
2014-03-24 10:53         ` Ian Campbell
2014-03-19 14:17 ` [RFC PATCH v1 02/10] xen/arm: register mmio handler at runtime vijay.kilari
2014-03-20 13:18   ` Julien Grall
2014-03-21 13:19     ` Andrii Tseglytskyi
2014-03-21 17:17   ` Ian Campbell
2014-03-21 17:23     ` Julien Grall
2014-03-26 12:29       ` Vijay Kilari
2014-03-26 14:47         ` Julien Grall
2014-03-27  5:40           ` Vijay Kilari
2014-03-27 15:02             ` Julien Grall
2014-04-01  9:34               ` Vijay Kilari
2014-04-01 11:00                 ` Julien Grall
2014-04-01 12:32                   ` Vijay Kilari
2014-04-01 12:44                     ` Ian Campbell
2014-04-01 12:51                     ` Julien Grall
2014-04-01 13:05                       ` Vijay Kilari
2014-04-01 13:56                         ` Julien Grall
2014-03-19 14:17 ` [RFC PATCH v1 03/10] xen/arm: move vgic data to vgic driver vijay.kilari
2014-03-20 13:51   ` Julien Grall
2014-03-21 17:23     ` Ian Campbell
2014-03-22  9:20       ` Vijay Kilari
2014-03-24 10:57         ` Ian Campbell
2014-03-26 11:44           ` Vijay Kilari
2014-03-26 12:00             ` Ian Campbell
2014-03-26 12:42               ` Vijay Kilari
2014-03-22  9:17     ` Vijay Kilari
2014-03-20 17:14   ` Stefano Stabellini
2014-03-20 17:56     ` Julien Grall
2014-03-20 18:11       ` Stefano Stabellini
2014-03-21  9:22         ` Ian Campbell
2014-03-19 14:17 ` [RFC PATCH v1 04/10] arm/xen: move gic save and restore registers to gic driver vijay.kilari
2014-03-20 15:22   ` Julien Grall
2014-03-21 17:26     ` Ian Campbell
2014-03-22  9:22     ` Vijay Kilari
2014-03-20 17:23   ` Stefano Stabellini
2014-03-21 17:28     ` Ian Campbell
2014-03-22  9:27     ` Vijay Kilari
2014-03-19 14:17 ` [RFC PATCH v1 05/10] xen/arm: move gic definitions to seperate file vijay.kilari
2014-03-20 15:13   ` Julien Grall
2014-03-19 14:17 ` [RFC PATCH v1 06/10] xen/arm: split gic driver into generic and gicv2 driver vijay.kilari
2014-03-20 11:55   ` Stefano Stabellini
2014-03-22  9:32     ` Vijay Kilari
2014-03-23 14:43       ` Julien Grall
2014-03-24 11:01         ` Ian Campbell
2014-03-20 16:02   ` Julien Grall
2014-03-21 17:32     ` Ian Campbell
2014-03-21 17:37       ` Julien Grall
2014-03-22  9:40     ` Vijay Kilari
2014-03-23 15:05       ` Julien Grall
2014-03-20 16:39   ` Stefano Stabellini
2014-03-21 17:38   ` Ian Campbell
2014-03-22  9:59     ` Vijay Kilari
2014-03-24 11:06       ` Ian Campbell
2014-03-19 14:17 ` [RFC PATCH v1 07/10] xen/arm: split vgic into generic and GIC v2 specific drivers vijay.kilari
2014-03-19 14:17 ` [RFC PATCH v1 08/10] xen/arm: Add support for GIC v3 vijay.kilari
2014-03-20 12:37   ` Stefano Stabellini
2014-03-22 10:07     ` Vijay Kilari
2014-03-24 11:28       ` Ian Campbell
2014-03-24 17:01       ` Stefano Stabellini
2014-03-26 13:16         ` Vijay Kilari
2014-03-26 17:22           ` Stefano Stabellini
2014-03-20 16:40   ` Julien Grall
2014-03-22 10:21     ` Vijay Kilari
2014-03-23 14:49       ` Julien Grall
2014-03-24 11:26         ` Ian Campbell
2014-03-24 11:50           ` Julien Grall
2014-03-24 17:02       ` Stefano Stabellini
2014-03-19 14:17 ` [RFC PATCH v1 09/10] xen/arm: Add vgic " vijay.kilari
2014-03-20 12:38   ` Stefano Stabellini
2014-03-19 14:17 ` [RFC PATCH v1 10/10] xen/arm: GICv3 device tree parsing vijay.kilari
2014-03-20 16:08   ` Julien Grall
2014-03-22 10:30     ` Vijay Kilari
2014-03-24 11:43       ` Ian Campbell
2014-03-24 12:03         ` Julien Grall
2014-03-24 12:07           ` Ian Campbell
2014-03-24 12:08           ` Julien Grall
2014-03-24 17:34             ` Stefano Stabellini
2014-03-24 18:00               ` Julien Grall
2014-03-25 11:04                 ` Stefano Stabellini
2014-03-25 12:33                   ` Julien Grall
2014-03-25 12:34                     ` Julien Grall
2014-04-01 12:59                     ` Ian Campbell
2014-04-01 13:07                       ` Julien Grall
2014-03-20 11:55 ` [RFC PATCH v1 00/10] xen/arm: Add GICv3 support Stefano Stabellini

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=532AE392.1080107@linaro.org \
    --to=julien.grall@linaro.org \
    --cc=Ian.Campbell@citrix.com \
    --cc=Prasun.Kapoor@caviumnetworks.com \
    --cc=stefano.stabellini@citrix.com \
    --cc=stefano.stabellini@eu.citrix.com \
    --cc=vijay.kilari@gmail.com \
    --cc=vijaya.kumar@caviumnetworks.com \
    --cc=xen-devel@lists.xen.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).