[<prev] [next>] [<thread-prev] [day] [month] [year] [list]
Message-ID: <ae888d-67c2-768-6985-e5cca038987e@linux.intel.com>
Date: Mon, 14 Aug 2023 18:11:24 +0300 (EEST)
From: Ilpo Järvinen <ilpo.jarvinen@...ux.intel.com>
To: "Peng Fan (OSS)" <peng.fan@....nxp.com>
cc: bhelgaas@...gle.com, pali@...nel.org, linux-pci@...r.kernel.org,
LKML <linux-kernel@...r.kernel.org>, Peng Fan <peng.fan@....com>
Subject: Re: [PATCH] pci: introduce static_nr to indicate domain_nr from
which IDA
On Sat, 12 Aug 2023, Peng Fan (OSS) wrote:
> From: Peng Fan <peng.fan@....com>
>
> When pci node is created used an overlay, when the overlay is
pci -> PCI
created used -> created using
is created -> was created ? The current form with "is created" + "is
reverted/destroyed" is contradictory.
Double "when" makes it hard to understand what you say. Perhaps convert
the second when to "and".
> reverted/destroyed, the "linux,pci-domain" property is no longer
> existed
property no longer exists
>, so of_get_pci_domain_nr will return failure. Then
> of_pci_bus_release_domain_nr will actually use the dynamic IDA,
> even if the IDA was allocated in static IDA.
>
> Introduce a static_nr field in pci_bus to indicate the IDA
> is in dynamic or static IDA to address the upper issue.
I'd say:
... to indicated whether the IDA is a dynamic or static in order to
free the correct one.
> Fixes: c14f7ccc9f5d ("PCI: Assign PCI domain IDs by ida_alloc()")
> Signed-off-by: Peng Fan <peng.fan@....com>
> ---
> drivers/pci/pci.c | 22 ++++++++++++++--------
> include/linux/pci.h | 1 +
> 2 files changed, 15 insertions(+), 8 deletions(-)
>
> diff --git a/drivers/pci/pci.c b/drivers/pci/pci.c
> index 60230da957e0..5c98502bcda6 100644
> --- a/drivers/pci/pci.c
> +++ b/drivers/pci/pci.c
> @@ -6881,10 +6881,10 @@ static void of_pci_reserve_static_domain_nr(void)
> }
> }
>
> -static int of_pci_bus_find_domain_nr(struct device *parent)
> +static int of_pci_bus_find_domain_nr(struct pci_bus *bus, struct device *parent)
> {
> static bool static_domains_reserved = false;
> - int domain_nr;
> + int domain_nr, ret;
>
> /* On the first call scan device tree for static allocations. */
> if (!static_domains_reserved) {
> @@ -6892,6 +6892,8 @@ static int of_pci_bus_find_domain_nr(struct device *parent)
> static_domains_reserved = true;
> }
>
> + bus->static_nr = 0;
> +
> if (parent) {
> /*
> * If domain is in DT, allocate it in static IDA. This
> @@ -6899,10 +6901,14 @@ static int of_pci_bus_find_domain_nr(struct device *parent)
> * in DT.
> */
> domain_nr = of_get_pci_domain_nr(parent->of_node);
> - if (domain_nr >= 0)
> - return ida_alloc_range(&pci_domain_nr_static_ida,
> - domain_nr, domain_nr,
> - GFP_KERNEL);
> + if (domain_nr >= 0) {
> + ret = ida_alloc_range(&pci_domain_nr_static_ida,
> + domain_nr, domain_nr, GFP_KERNEL);
> + if (ret >= 0)
> + bus->static_nr = 1;
> +
> + return ret;
> + }
> }
>
> /*
> @@ -6920,7 +6926,7 @@ static void of_pci_bus_release_domain_nr(struct pci_bus *bus, struct device *par
> return;
>
> /* Release domain from IDA where it was allocated. */
> - if (of_get_pci_domain_nr(parent->of_node) == bus->domain_nr)
> + if (bus->static_nr)
> ida_free(&pci_domain_nr_static_ida, bus->domain_nr);
> else
> ida_free(&pci_domain_nr_dynamic_ida, bus->domain_nr);
> @@ -6928,7 +6934,7 @@ static void of_pci_bus_release_domain_nr(struct pci_bus *bus, struct device *par
>
> int pci_bus_find_domain_nr(struct pci_bus *bus, struct device *parent)
> {
> - return acpi_disabled ? of_pci_bus_find_domain_nr(parent) :
> + return acpi_disabled ? of_pci_bus_find_domain_nr(bus, parent) :
> acpi_pci_bus_find_domain_nr(bus);
> }
>
> diff --git a/include/linux/pci.h b/include/linux/pci.h
> index eeb2e6f6130f..6dd16e069ab8 100644
> --- a/include/linux/pci.h
> +++ b/include/linux/pci.h
> @@ -665,6 +665,7 @@ struct pci_bus {
> unsigned char cur_bus_speed; /* enum pci_bus_speed */
> #ifdef CONFIG_PCI_DOMAINS_GENERIC
> int domain_nr;
> + unsigned int static_nr:1;
> #endif
>
> char name[48];
If you put it into where the other unsigned int xx:1's are in the struct,
it won't increase the size of the struct (but the downside is that they're
no longer next to each other).
--
i.
Powered by blists - more mailing lists