Re: [PATCH V5 7/7] acpi, gicv3, its: Use MADT ITS subtable to do PCI/MSI domain initialization.

[Marc Zyngier <marc.zyngier@arm.com>]

On Tue, 31 May 2016 13:19:44 +0200
Tomasz Nowicki <tn@semihalf.com> wrote:

> Let ACPI build ITS PCI MSI domain. ACPI is responsible for retrieving
> inner domain token and passing it on to its_pci_msi_init_one generic
> init call.
> 
> We have now full PCI MSI domain stack, thus we can enable ITS initialization
> from GICv3 core driver for ACPI scenario.
> 
> Signed-off-by: Tomasz Nowicki <tn@semihalf.com>
> ---
>  drivers/irqchip/irq-gic-v3-its-pci-msi.c | 42 +++++++++++++++++++++++++++++++-
>  drivers/irqchip/irq-gic-v3.c             |  3 +--
>  2 files changed, 42 insertions(+), 3 deletions(-)
> 
> diff --git a/drivers/irqchip/irq-gic-v3-its-pci-msi.c b/drivers/irqchip/irq-gic-v3-its-pci-msi.c
> index 23679ae..985f5d9 100644
> --- a/drivers/irqchip/irq-gic-v3-its-pci-msi.c
> +++ b/drivers/irqchip/irq-gic-v3-its-pci-msi.c
> @@ -15,6 +15,7 @@
>   * along with this program.  If not, see <http://www.gnu.org/licenses/>.
>   */
>  
> +#include <linux/iort.h>
>  #include <linux/msi.h>
>  #include <linux/of.h>
>  #include <linux/of_irq.h>
> @@ -144,10 +145,49 @@ static int __init its_pci_of_msi_init(void)
>  	return 0;
>  }
>  
> +#ifdef CONFIG_ACPI
> +
> +static int __init
> +its_pci_msi_parse_madt(struct acpi_subtable_header *header,
> +		       const unsigned long end)
> +{
> +	struct acpi_madt_generic_translator *its_entry;
> +	struct fwnode_handle *dom_handle;
> +	char node_name[6 + 16 + 1]; // ITS identification name

Jeezus... Where is that coming from? Is that the size of "ITS@0x"+16
hex digits and a final NUL? For the sake of keeping things
maintainable/readable, please use kasprintf/kfree.

> +
> +	its_entry = (struct acpi_madt_generic_translator *)header;
> +	snprintf(node_name, 23, "ITS@0x%lx", (long)its_entry->base_address);
> +	dom_handle = iort_its_find_domain_token(its_entry->translation_id);
> +	if (!dom_handle) {
> +		pr_err("%s: Unable to locate ITS domain handle\n", node_name);
> +		return 0;

return 0? not even a slap on the wrist? Hum...

> +	}
> +
> +	if (its_pci_msi_init_one(dom_handle, node_name))
> +		return 0;

Same here.

> +
> +	pr_info("PCI/MSI: %s domain created\n", node_name);
> +	return 0;
> +}
> +
> +static int __init its_pci_acpi_msi_init(void)
> +{
> +	acpi_table_parse_madt(ACPI_MADT_TYPE_GENERIC_TRANSLATOR,
> +			      its_pci_msi_parse_madt, 0);
> +	return 0;
> +}
> +#else
> +inline static int __init its_pci_acpi_msi_init(void)

static only.

> +{
> +	return 0;
> +}
> +#endif
> +
>  static int __init its_pci_msi_init(void)
>  {
>  	its_pci_of_msi_init();
> +	its_pci_acpi_msi_init();
> +
>  	return 0;
>  }
> -
>  early_initcall(its_pci_msi_init);
> diff --git a/drivers/irqchip/irq-gic-v3.c b/drivers/irqchip/irq-gic-v3.c
> index cee8296..41c1cda 100644
> --- a/drivers/irqchip/irq-gic-v3.c
> +++ b/drivers/irqchip/irq-gic-v3.c
> @@ -951,8 +951,7 @@ static int __init gic_init_bases(void __iomem *dist_base,
>  
>  	set_handle_irq(gic_handle_irq);
>  
> -	if (IS_ENABLED(CONFIG_ARM_GIC_V3_ITS) && gic_dist_supports_lpis() &&
> -	    to_of_node(handle)) /* Temp hack to prevent ITS init for ACPI */
> +	if (IS_ENABLED(CONFIG_ARM_GIC_V3_ITS) && gic_dist_supports_lpis())
>  		its_init(handle, &gic_data.rdists, gic_data.domain);
>  
>  	gic_smp_init();

Thanks,

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