Date:   Thu, 25 Nov 2021 14:03:36 +0000
Message-ID: <87a6hscoxz.wl-maz@kernel.org>
From:   Marc Zyngier <maz@kernel.org>
To:     Pali =?UTF-8?B?Um9ow6Fy?= <pali@kernel.org>
Cc:     Andrew Lunn <andrew@lunn.ch>,
        Gregory Clement <gregory.clement@bootlin.com>,
        Sebastian Hesselbarth <sebastian.hesselbarth@gmail.com>,
        Thomas Gleixner <tglx@linutronix.de>,
        Marek =?UTF-8?B?QmVow7pu?= <kabel@kernel.org>,
        linux-arm-kernel@lists.infradead.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH 2/2] irqchip/armada-370-xp: Fix support for Multi-MSI interrupts
In-Reply-To: <20211125130057.26705-2-pali@kernel.org>
References: <20211125130057.26705-1-pali@kernel.org>
        <20211125130057.26705-2-pali@kernel.org>
User-Agent: Wanderlust/2.15.9 (Almost Unreal) SEMI-EPG/1.14.7 (Harue)
 FLIM-LB/1.14.9 (=?UTF-8?B?R29qxY0=?=) APEL-LB/10.8 EasyPG/1.0.0 Emacs/27.1
 (x86_64-pc-linux-gnu) MULE/6.0 (HANACHIRUSATO)
MIME-Version: 1.0 (generated by SEMI-EPG 1.14.7 - "Harue")
Content-Type: text/plain; charset=UTF-8
Content-Transfer-Encoding: 8BIT
Precedence: bulk

On Thu, 25 Nov 2021 13:00:57 +0000,
Pali Rohár <pali@kernel.org> wrote:
> 
> irq-armada-370-xp driver already sets MSI_FLAG_MULTI_PCI_MSI flag into
> msi_domain_info structure. But allocated interrupt numbers for Multi-MSI
> needs to be properly aligned otherwise devices send MSI interrupt with
> wrong number.
> 
> Fix this issue by using function bitmap_find_free_region() instead of
> bitmap_find_next_zero_area() to allocate aligned interrupt numbers.
> 
> Signed-off-by: Pali Rohár <pali@kernel.org>
> Fixes: a71b9412c90c ("irqchip/armada-370-xp: Allow allocation of multiple MSIs")
> Cc: stable@vger.kernel.org
> ---
>  drivers/irqchip/irq-armada-370-xp.c | 14 +++++---------
>  1 file changed, 5 insertions(+), 9 deletions(-)
> 
> diff --git a/drivers/irqchip/irq-armada-370-xp.c b/drivers/irqchip/irq-armada-370-xp.c
> index 41ad745cf343..5b8d571c041d 100644
> --- a/drivers/irqchip/irq-armada-370-xp.c
> +++ b/drivers/irqchip/irq-armada-370-xp.c
> @@ -232,16 +232,12 @@ static int armada_370_xp_msi_alloc(struct irq_domain *domain, unsigned int virq,
>  	int hwirq, i;
>  
>  	mutex_lock(&msi_used_lock);
> +	hwirq = bitmap_find_free_region(msi_used, PCI_MSI_DOORBELL_NR,
> +					order_base_2(nr_irqs));
> +	mutex_unlock(&msi_used_lock);
>  
> -	hwirq = bitmap_find_next_zero_area(msi_used, PCI_MSI_DOORBELL_NR,
> -					   0, nr_irqs, 0);
> -	if (hwirq >= PCI_MSI_DOORBELL_NR) {
> -		mutex_unlock(&msi_used_lock);
> +	if (hwirq < 0)
>  		return -ENOSPC;
> -	}
> -
> -	bitmap_set(msi_used, hwirq, nr_irqs);
> -	mutex_unlock(&msi_used_lock);
>  
>  	for (i = 0; i < nr_irqs; i++) {
>  		irq_domain_set_info(domain, virq + i, hwirq + i,
> @@ -259,7 +255,7 @@ static void armada_370_xp_msi_free(struct irq_domain *domain,
>  	struct irq_data *d = irq_domain_get_irq_data(domain, virq);
>  
>  	mutex_lock(&msi_used_lock);
> -	bitmap_clear(msi_used, d->hwirq, nr_irqs);
> +	bitmap_release_region(msi_used, d->hwirq, order_base_2(nr_irqs));
>  	mutex_unlock(&msi_used_lock);
>  }
>  
> -- 
> 2.20.1
> 
> 

Acked-by: Marc Zyngier <maz@kernel.org>

	M.

-- 
Without deviation from the norm, progress is not possible.