[PATCH v5] PCI: ARM: add support for generic PCI host controller
Liviu Dudau
Liviu.Dudau at arm.com
Wed Mar 5 13:10:57 EST 2014
On Mon, Mar 03, 2014 at 06:46:40PM +0000, Will Deacon wrote:
> This patch adds support for a generic PCI host controller, such as a
> firmware-initialised device with static windows or an emulation by
> something such as kvmtool.
>
> The controller itself has no configuration registers and has its address
> spaces described entirely by the device-tree (using the bindings from
> ePAPR). Both CAM and ECAM are supported for Config Space accesses.
>
> Corresponding documentation is added for the DT binding.
>
> Signed-off-by: Will Deacon <will.deacon at arm.com>
> ---
>
> Hi,
>
> This is version 5 of the patches previously posted here:
>
> v1: http://lists.infradead.org/pipermail/linux-arm-kernel/2014-February/229679.html
> v2: http://lists.infradead.org/pipermail/linux-arm-kernel/2014-February/232213.html
> v3: http://lists.infradead.org/pipermail/linux-arm-kernel/2014-February/233491.html
> v4: http://lists.infradead.org/pipermail/linux-arm-kernel/2014-February/235468.html
>
> Changes since v4 include:
>
> - Changed binding example to use I/O bus address 0x0
>
> - Fixed and simplified rounding code for I/O ranges so that we don't
> mess around with the registered resource
>
> - Finally fixed I/O offset handling
>
> Note that I didn't use of_property_read_bool for parsing "linux,pci-probe-only"
> since that would interfere with "pci=firmware" being passed on the command
> line (meaning we have a tri-state). I also didn't move the resource
> registration into gen_pci_calc_*_offset, since that duplicates code and
> complicates the freeing paths.
>
> As per usual, all feedback welcome. The next step is probably to work out
> a set of library functions that can be extracted from this driver.
>
> Will
>
> .../devicetree/bindings/pci/host-generic-pci.txt | 100 ++++++
> drivers/pci/host/Kconfig | 7 +
> drivers/pci/host/Makefile | 1 +
> drivers/pci/host/pci-host-generic.c | 369 +++++++++++++++++++++
> 4 files changed, 477 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/pci/host-generic-pci.txt
> create mode 100644 drivers/pci/host/pci-host-generic.c
>
> diff --git a/Documentation/devicetree/bindings/pci/host-generic-pci.txt b/Documentation/devicetree/bindings/pci/host-generic-pci.txt
> new file mode 100644
> index 000000000000..98788c153302
> --- /dev/null
> +++ b/Documentation/devicetree/bindings/pci/host-generic-pci.txt
> @@ -0,0 +1,100 @@
> +* Generic PCI host controller
> +
> +Firmware-initialised PCI host controllers and PCI emulations, such as the
> +virtio-pci implementations found in kvmtool and other para-virtualised
> +systems, do not require driver support for complexities such as regulator
> +and clock management. In fact, the controller may not even require the
> +configuration of a control interface by the operating system, instead
> +presenting a set of fixed windows describing a subset of IO, Memory and
> +Configuration Spaces.
> +
> +Such a controller can be described purely in terms of the standardized device
> +tree bindings communicated in pci.txt:
> +
> +
> +Properties of the host controller node:
> +
> +- compatible : Must be "pci-host-cam-generic" or "pci-host-ecam-generic"
> + depending on the layout of configuration space (CAM vs
> + ECAM respectively).
> +
> +- device_type : Must be "pci".
> +
> +- ranges : As described in IEEE Std 1275-1994, but must provide
> + at least a definition of non-prefetchable memory. One
> + or both of prefetchable Memory and IO Space may also
> + be provided.
> +
> +- bus-range : Optional property (also described in IEEE Std 1275-1994)
> + to indicate the range of bus numbers for this controller.
> + If absent, defaults to <0 255> (i.e. all buses).
> +
> +- #address-cells : Must be 3.
> +
> +- #size-cells : Must be 2.
> +
> +- reg : The Configuration Space base address and size, as accessed
> + from the parent bus.
> +
> +
> +Properties of the /chosen node:
> +
> +- linux,pci-probe-only
> + : Optional property which takes a single-cell argument.
> + If '0', then Linux will assign devices in its usual manner,
> + otherwise it will not try to assign devices and instead use
> + them as they are configured already.
> +
> +Configuration Space is assumed to be memory-mapped (as opposed to being
> +accessed via an ioport) and laid out with a direct correspondence to the
> +geography of a PCI bus address by concatenating the various components to
> +form an offset.
> +
> +For CAM, this 24-bit offset is:
> +
> + cfg_offset(bus, device, function, register) =
> + bus << 16 | device << 11 | function << 8 | register
> +
> +Whilst ECAM extends this by 4 bits to accomodate 4k of function space:
> +
> + cfg_offset(bus, device, function, register) =
> + bus << 20 | device << 15 | function << 12 | register
> +
> +Interrupt mapping is exactly as described in `Open Firmware Recommended
> +Practice: Interrupt Mapping' and requires the following properties:
> +
> +- #interrupt-cells : Must be 1
> +
> +- interrupt-map : <see aforementioned specification>
> +
> +- interrupt-map-mask : <see aforementioned specification>
> +
> +
> +Example:
> +
> +pci {
> + compatible = "pci-host-cam-generic"
> + device_type = "pci";
> + #address-cells = <3>;
> + #size-cells = <2>;
> + bus-range = <0x0 0x1>;
> +
> + // CPU_PHYSICAL(2) SIZE(2)
> + reg = <0x0 0x40000000 0x0 0x1000000>;
> +
> + // BUS_ADDRESS(3) CPU_PHYSICAL(2) SIZE(2)
> + ranges = <0x0000000 0x0 0x01000000 0x0 0x01000000 0x0 0x00010000>,
> + <0x1000000 0x0 0x41000000 0x0 0x41000000 0x0 0x3f000000>;
> +
> +
> + #interrupt-cells = <0x1>;
> +
> + // PCI_DEVICE(3) INT#(1) CONTROLLER(PHANDLE) CONTROLLER_DATA(3)
> + interrupt-map = < 0x0 0x0 0x0 0x1 &gic 0x0 0x4 0x1
> + 0x800 0x0 0x0 0x1 &gic 0x0 0x5 0x1
> + 0x1000 0x0 0x0 0x1 &gic 0x0 0x6 0x1
> + 0x1800 0x0 0x0 0x1 &gic 0x0 0x7 0x1>;
> +
> + // PCI_DEVICE(3) INT#(1)
> + interrupt-map-mask = <0xf800 0x0 0x0 0x7>;
> +}
> diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig
> index 47d46c6d8468..f91637d47065 100644
> --- a/drivers/pci/host/Kconfig
> +++ b/drivers/pci/host/Kconfig
> @@ -33,4 +33,11 @@ config PCI_RCAR_GEN2
> There are 3 internal PCI controllers available with a single
> built-in EHCI/OHCI host controller present on each one.
>
> +config PCI_HOST_GENERIC
> + bool "Generic PCI host controller"
> + depends on ARM && OF
> + help
> + Say Y here if you want to support a simple generic PCI host
> + controller, such as the one emulated by kvmtool.
> +
> endmenu
> diff --git a/drivers/pci/host/Makefile b/drivers/pci/host/Makefile
> index 13fb3333aa05..bd1bf1ab4ac8 100644
> --- a/drivers/pci/host/Makefile
> +++ b/drivers/pci/host/Makefile
> @@ -4,3 +4,4 @@ obj-$(CONFIG_PCI_IMX6) += pci-imx6.o
> obj-$(CONFIG_PCI_MVEBU) += pci-mvebu.o
> obj-$(CONFIG_PCI_TEGRA) += pci-tegra.o
> obj-$(CONFIG_PCI_RCAR_GEN2) += pci-rcar-gen2.o
> +obj-$(CONFIG_PCI_HOST_GENERIC) += pci-host-generic.o
> diff --git a/drivers/pci/host/pci-host-generic.c b/drivers/pci/host/pci-host-generic.c
> new file mode 100644
> index 000000000000..b761f3e64a6a
> --- /dev/null
> +++ b/drivers/pci/host/pci-host-generic.c
> @@ -0,0 +1,369 @@
> +/*
> + * Simple, generic PCI host controller driver targetting firmware-initialised
> + * systems and virtual machines (e.g. the PCI emulation provided by kvmtool).
> + *
> + * This program is free software; you can redistribute it and/or modify
> + * it under the terms of the GNU General Public License version 2 as
> + * published by the Free Software Foundation.
> + *
> + * This program is distributed in the hope that it will be useful,
> + * but WITHOUT ANY WARRANTY; without even the implied warranty of
> + * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
> + * GNU General Public License for more details.
> + *
> + * You should have received a copy of the GNU General Public License
> + * along with this program. If not, see <http://www.gnu.org/licenses/>.
> + *
> + * Copyright (C) 2014 ARM Limited
> + *
> + * Author: Will Deacon <will.deacon at arm.com>
> + */
> +
> +#include <linux/kernel.h>
> +#include <linux/module.h>
> +#include <linux/of_address.h>
> +#include <linux/of_pci.h>
> +#include <linux/platform_device.h>
> +
> +struct gen_pci_cfg_bus_ops {
> + u32 bus_shift;
> + void __iomem *(*map_bus)(struct pci_bus *, unsigned int, int);
> +};
> +
> +struct gen_pci_cfg_windows {
> + struct resource res;
> + struct resource bus_range;
> + void __iomem **win;
> +
> + const struct gen_pci_cfg_bus_ops *ops;
> +};
> +
> +struct gen_pci {
> + struct pci_host_bridge host;
> + struct gen_pci_cfg_windows cfg;
> +};
> +
> +static void __iomem *gen_pci_map_cfg_bus_cam(struct pci_bus *bus,
> + unsigned int devfn,
> + int where)
> +{
> + struct pci_sys_data *sys = bus->sysdata;
> + struct gen_pci *pci = sys->private_data;
> + resource_size_t idx = bus->number - pci->cfg.bus_range.start;
> +
> + return pci->cfg.win[idx] + ((devfn << 8) | where);
> +}
> +
> +static struct gen_pci_cfg_bus_ops gen_pci_cfg_cam_bus_ops = {
> + .bus_shift = 16,
> + .map_bus = gen_pci_map_cfg_bus_cam,
> +};
> +
> +static void __iomem *gen_pci_map_cfg_bus_ecam(struct pci_bus *bus,
> + unsigned int devfn,
> + int where)
> +{
> + struct pci_sys_data *sys = bus->sysdata;
> + struct gen_pci *pci = sys->private_data;
> + resource_size_t idx = bus->number - pci->cfg.bus_range.start;
> +
> + return pci->cfg.win[idx] + ((devfn << 12) | where);
> +}
> +
> +static struct gen_pci_cfg_bus_ops gen_pci_cfg_ecam_bus_ops = {
> + .bus_shift = 20,
> + .map_bus = gen_pci_map_cfg_bus_ecam,
> +};
> +
> +static int gen_pci_config_read(struct pci_bus *bus, unsigned int devfn,
> + int where, int size, u32 *val)
> +{
> + void __iomem *addr;
> + struct pci_sys_data *sys = bus->sysdata;
> + struct gen_pci *pci = sys->private_data;
> +
> + addr = pci->cfg.ops->map_bus(bus, devfn, where);
> +
> + switch (size) {
> + case 1:
> + *val = readb(addr);
> + break;
> + case 2:
> + *val = readw(addr);
> + break;
> + default:
> + *val = readl(addr);
> + }
> +
> + return PCIBIOS_SUCCESSFUL;
> +}
> +
> +static int gen_pci_config_write(struct pci_bus *bus, unsigned int devfn,
> + int where, int size, u32 val)
> +{
> + void __iomem *addr;
> + struct pci_sys_data *sys = bus->sysdata;
> + struct gen_pci *pci = sys->private_data;
> +
> + addr = pci->cfg.ops->map_bus(bus, devfn, where);
> +
> + switch (size) {
> + case 1:
> + writeb(val, addr);
> + break;
> + case 2:
> + writew(val, addr);
> + break;
> + default:
> + writel(val, addr);
> + }
> +
> + return PCIBIOS_SUCCESSFUL;
> +}
> +
> +static struct pci_ops gen_pci_ops = {
> + .read = gen_pci_config_read,
> + .write = gen_pci_config_write,
> +};
> +
> +static int gen_pci_calc_io_offset(struct device *dev,
> + struct of_pci_range *range,
> + struct resource *res,
> + resource_size_t *offset)
> +{
> + static atomic_t wins = ATOMIC_INIT(0);
> + int err, idx, max_win;
> + unsigned int window;
> +
> + if (!PAGE_ALIGNED(range->cpu_addr))
> + return -EINVAL;
> +
> + max_win = (IO_SPACE_LIMIT + 1) / SZ_64K;
> + idx = atomic_inc_return(&wins);
> + if (idx >= max_win)
> + return -ENOSPC;
> +
> + window = (idx - 1) * SZ_64K;
> + err = pci_ioremap_io(window, range->cpu_addr);
> + if (err)
> + return err;
> +
> + of_pci_range_to_resource(range, dev->of_node, res);
> + res->start = window;
> + res->end = res->start + range->size - 1;
> + *offset = window - range->pci_addr;
If my fix for of_pci_range_to_resource will get accepted, this
function will have to morph into the pci_register_io_range() call.
Would that cause problems for other users of of_pci_range_to_resource()
(namely mach-integrator ?)
Otherwise, looks good to me. You can add my (worthless?)
Reviewed-by: Liviu Dudau <Liviu.Dudau at arm.com>
Cheers,
Liviu
> + return 0;
> +}
> +
> +static int gen_pci_calc_mem_offset(struct device *dev,
> + struct of_pci_range *range,
> + struct resource *res,
> + resource_size_t *offset)
> +{
> + of_pci_range_to_resource(range, dev->of_node, res);
> + *offset = range->cpu_addr - range->pci_addr;
> + return 0;
> +}
> +
> +static const struct of_device_id gen_pci_of_match[] = {
> + { .compatible = "pci-host-cam-generic",
> + .data = &gen_pci_cfg_cam_bus_ops },
> +
> + { .compatible = "pci-host-ecam-generic",
> + .data = &gen_pci_cfg_ecam_bus_ops },
> +
> + { },
> +};
> +MODULE_DEVICE_TABLE(of, gen_pci_of_match);
> +
> +static int gen_pci_setup(int nr, struct pci_sys_data *sys)
> +{
> + int err, res_valid;
> + u8 bus_max;
> + struct pci_host_bridge_window *win;
> + struct resource *bus_range;
> + resource_size_t busn;
> + const char *type;
> + struct of_pci_range range;
> + struct of_pci_range_parser parser;
> + const struct of_device_id *of_id;
> + const int *prop;
> + struct gen_pci *pci = sys->private_data;
> + struct device *dev = pci->host.dev.parent;
> + struct device_node *np = dev->of_node;
> +
> + type = of_get_property(np, "device_type", NULL);
> + if (!type || strcmp(type, "pci")) {
> + dev_err(dev, "invalid \"device_type\" %s\n", type);
> + return -EINVAL;
> + }
> +
> + if (of_pci_range_parser_init(&parser, np)) {
> + dev_err(dev, "missing \"ranges\" property\n");
> + return -EINVAL;
> + }
> +
> + prop = of_get_property(of_chosen, "linux,pci-probe-only", NULL);
> + if (prop) {
> + if (*prop)
> + pci_add_flags(PCI_PROBE_ONLY);
> + else
> + pci_clear_flags(PCI_PROBE_ONLY);
> + }
> +
> + if (of_pci_parse_bus_range(np, &pci->cfg.bus_range))
> + pci->cfg.bus_range = (struct resource) {
> + .name = np->name,
> + .start = 0,
> + .end = 0xff,
> + .flags = IORESOURCE_BUS,
> + };
> +
> + err = of_address_to_resource(np, 0, &pci->cfg.res);
> + if (err) {
> + dev_err(dev, "missing \"reg\" property\n");
> + return err;
> + }
> +
> + pci->cfg.win = devm_kcalloc(dev, resource_size(&pci->cfg.bus_range),
> + sizeof(*pci->cfg.win), GFP_KERNEL);
> + if (!pci->cfg.win)
> + return -ENOMEM;
> +
> + of_id = of_match_node(gen_pci_of_match, np);
> + pci->cfg.ops = of_id->data;
> + INIT_LIST_HEAD(&pci->host.windows);
> +
> + /* Limit the bus-range to fit within reg */
> + bus_max = pci->cfg.bus_range.start +
> + (resource_size(&pci->cfg.res) >> pci->cfg.ops->bus_shift) - 1;
> + pci->cfg.bus_range.end = min_t(resource_size_t, pci->cfg.bus_range.end,
> + bus_max);
> +
> + /* Create and register resources from the ranges property */
> + res_valid = 0;
> + for_each_of_pci_range(&parser, &range) {
> + struct resource *parent, *res;
> + resource_size_t offset;
> + u32 restype = range.flags & IORESOURCE_TYPE_BITS;
> +
> + res = devm_kmalloc(dev, sizeof(*res), GFP_KERNEL);
> + if (!res) {
> + err = -ENOMEM;
> + goto out_release_res;
> + }
> +
> + switch (restype) {
> + case IORESOURCE_IO:
> + parent = &ioport_resource;
> + err = gen_pci_calc_io_offset(dev, &range, res, &offset);
> + break;
> + case IORESOURCE_MEM:
> + parent = &iomem_resource;
> + err = gen_pci_calc_mem_offset(dev, &range, res, &offset);
> + res_valid |= !(res->flags & IORESOURCE_PREFETCH || err);
> + break;
> + default:
> + err = -EINVAL;
> + continue;
> + }
> +
> + if (err) {
> + dev_warn(dev,
> + "error %d: failed to add resource [type 0x%x, %lld bytes]\n",
> + err, restype, range.size);
> + continue;
> + }
> +
> + err = request_resource(parent, res);
> + if (err)
> + goto out_release_res;
> +
> + pci_add_resource_offset(&sys->resources, res, offset);
> + }
> +
> + if (!res_valid) {
> + dev_err(dev, "non-prefetchable memory resource required\n");
> + err = -EINVAL;
> + goto out_release_res;
> + }
> +
> + /* Map our Configuration Space windows */
> + if (!request_mem_region(pci->cfg.res.start,
> + resource_size(&pci->cfg.res),
> + "Configuration Space"))
> + goto out_release_res;
> +
> + bus_range = &pci->cfg.bus_range;
> + for (busn = bus_range->start; busn <= bus_range->end; ++busn) {
> + u32 idx = busn - bus_range->start;
> + u32 sz = 1 << pci->cfg.ops->bus_shift;
> +
> + pci->cfg.win[idx] = devm_ioremap(dev,
> + pci->cfg.res.start + busn * sz,
> + sz);
> + if (!pci->cfg.win[idx]) {
> + err = -ENOMEM;
> + goto out_unmap_cfg;
> + }
> + }
> +
> + /* Register bus resource */
> + pci_add_resource(&sys->resources, bus_range);
> + return 1;
> +
> +out_unmap_cfg:
> + while (busn-- > bus_range->start)
> + devm_iounmap(dev, pci->cfg.win[busn - bus_range->start]);
> +
> +out_release_res:
> + release_child_resources(&iomem_resource);
> + release_child_resources(&ioport_resource);
> +
> + list_for_each_entry(win, &sys->resources, list)
> + devm_kfree(dev, win->res);
> + pci_free_resource_list(&sys->resources);
> +
> + devm_kfree(dev, pci->cfg.win);
> + return err;
> +}
> +
> +static int gen_pci_probe(struct platform_device *pdev)
> +{
> + struct hw_pci hw;
> + struct gen_pci *pci;
> + struct device *dev = &pdev->dev;
> +
> + if (!dev->of_node)
> + return -ENODEV;
> +
> + pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
> + if (!pci)
> + return -ENOMEM;
> +
> + pci->host.dev.parent = dev;
> + hw = (struct hw_pci) {
> + .nr_controllers = 1,
> + .private_data = (void **)&pci,
> + .setup = gen_pci_setup,
> + .map_irq = of_irq_parse_and_map_pci,
> + .ops = &gen_pci_ops,
> + };
> +
> + pci_common_init_dev(dev, &hw);
> + return 0;
> +}
> +
> +static struct platform_driver gen_pci_driver = {
> + .driver = {
> + .name = "pci-host-generic",
> + .owner = THIS_MODULE,
> + .of_match_table = gen_pci_of_match,
> + },
> + .probe = gen_pci_probe,
> +};
> +module_platform_driver(gen_pci_driver);
> +
> +MODULE_DESCRIPTION("Generic PCI host driver");
> +MODULE_AUTHOR("Will Deacon <will.deacon at arm.com>");
> +MODULE_LICENSE("GPLv2");
> --
> 1.8.2.2
>
> --
> To unsubscribe from this list: send the line "unsubscribe linux-pci" in
> the body of a message to majordomo at vger.kernel.org
> More majordomo info at http://vger.kernel.org/majordomo-info.html
>
--
====================
| I would like to |
| fix the world, |
| but they're not |
| giving me the |
\ source code! /
---------------
¯\_(ツ)_/¯
More information about the linux-arm-kernel
mailing list