[PATCH RFC 1/1] irqchip/GICv2m: Add support for multiple v2m frames

Marc Zyngier marc.zyngier at arm.com
Fri Oct 9 02:10:27 PDT 2015


Hi Duc,

On 08/10/15 08:48, Duc Dang wrote:
> GICv2m driver currently only supports single v2m frame. This
> patch extend this driver to support multiple v2m frames. All of
> the v2m frames will be own by a single MSI domain. Each PCIe node
> can specify msi-parent as the first frame of the v2m frame list
> to be able to use all available v2m frames for MSI interrupts.
> 
> This patch should be applied on top of patch 
> (irqchip/GICv2m: Add workaround for APM X-Gene GICv2m erratum):
> https://lkml.org/lkml/2015/10/6/922
> 
> Signed-off-by: Duc Dang <dhdang at apm.com>
> ---
>  drivers/irqchip/irq-gic-v2m.c | 221 ++++++++++++++++++++++++++++++------------
>  1 file changed, 159 insertions(+), 62 deletions(-)
> 
> diff --git a/drivers/irqchip/irq-gic-v2m.c b/drivers/irqchip/irq-gic-v2m.c
> index 4c17c18..8ecaf9e 100644
> --- a/drivers/irqchip/irq-gic-v2m.c
> +++ b/drivers/irqchip/irq-gic-v2m.c
> @@ -51,13 +51,19 @@
>  #define GICV2M_NEEDS_SPI_OFFSET		0x00000001
>  
>  struct v2m_data {
> -	spinlock_t msi_cnt_lock;
>  	struct resource res;	/* GICv2m resource */
>  	void __iomem *base;	/* GICv2m virt address */
>  	u32 spi_start;		/* The SPI number that MSIs start */
>  	u32 nr_spis;		/* The number of SPIs for MSIs */
> -	unsigned long *bm;	/* MSI vector bitmap */
>  	u32 flags;		/* v2m flags for specific implementation */
> +	struct list_head list;  /* Link to other v2m frames */
> +};
> +
> +struct gicv2m_ctrlr {
> +	spinlock_t v2m_ctrlr_lock;      /* Lock for all v2m frames */
> +	u32 nr_vecs;                    /* Total MSI vectors */
> +	unsigned long *bm;              /* MSI vector bitmap */
> +	struct list_head v2m_frms;      /* List of v2m frames */
>  };
>  
>  static void gicv2m_mask_msi_irq(struct irq_data *d)
> @@ -98,11 +104,29 @@ static int gicv2m_set_affinity(struct irq_data *irq_data,
>  	return ret;
>  }
>  
> +static struct v2m_data *irq_data_to_v2m_frm(struct irq_data *data,
> +					    struct gicv2m_ctrlr *v2m_ctrlr)
> +{
> +	struct v2m_data *v2m;
> +
> +	list_for_each_entry(v2m, &v2m_ctrlr->v2m_frms, list) {
> +		if ((data->hwirq >= v2m->spi_start) &&
> +		    (data->hwirq < (v2m->spi_start + v2m->nr_spis)))
> +			return v2m;
> +	}
> +
> +	return NULL;
> +}
> +
>  static void gicv2m_compose_msi_msg(struct irq_data *data, struct msi_msg *msg)
>  {
> -	struct v2m_data *v2m = irq_data_get_irq_chip_data(data);
> -	phys_addr_t addr = v2m->res.start + V2M_MSI_SETSPI_NS;
> +	struct gicv2m_ctrlr *v2m_ctrlr = irq_data_get_irq_chip_data(data);
> +	struct v2m_data *v2m;
> +	phys_addr_t addr;
>  
> +	v2m = irq_data_to_v2m_frm(data, v2m_ctrlr);
> +	WARN_ON(!v2m);
> +	addr = v2m->res.start + V2M_MSI_SETSPI_NS;

At that particular point, I've stopped reading.

Pointlessly iterating over the frames when you can store the right one
in irq_data was bad enough, but having a WARN_ON() followed by a panic
is just not acceptable. Either you warn and avoid the crash, or you just
assume that the whole thing is sane and will never be inconsistent.

You are also changing the way this driver works, and not for the best.
I've just written something fairly similar, with the following diffstat:

 1 file changed, 76 insertions(+), 45 deletions(-)

The current infrastructure is sound, you just have to make use of it.

I've pushed the following (untested) patch:

https://git.kernel.org/cgit/linux/kernel/git/maz/arm-platforms.git/commit/?h=irq/multi-v2m&id=45e35adb60087792626570ff21bb23ab0f67a6ac

Let me know if that works for you.

Thanks,

	M.
-- 
Jazz is not dead. It just smells funny...



More information about the linux-arm-kernel mailing list