* [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe
@ 2016-01-08 23:59 David Daney
2016-01-08 23:59 ` [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers David Daney
` (2 more replies)
0 siblings, 3 replies; 6+ messages in thread
From: David Daney @ 2016-01-08 23:59 UTC (permalink / raw)
To: linux-arm-kernel
From: David Daney <david.daney@cavium.com>
Some Cavium ThunderX processors require quirky access methods for the
config space of the PCIe bridge.
There are two patches:
1) Refactor code in pci-host-generic so that it can more easily be
used by other drivers. This splits the driver for CAM and ECAM
access methods to a separate file from the common host driver code.
2) Add the ThunderX PCIe driver, which leverages the code in
pci-host-generic
Changes from v2: Improve device tree binding example as noted by Rob
Herring. Rename pcie-thunder-pem.* to pci-thunder-pem.* for better
consistency. Update MAINTAINERS to reflect the changes.
Changes from v1: Split CAM and ECAM code from common driver code as
suggested by Arnd Bergmann. Fix spelling errors in
pcie-thunder-pem.txt
David Daney (2):
PCI: generic: Refactor code to enable reuse by other drivers.
pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors.
.../devicetree/bindings/pci/pci-thunder-pem.txt | 43 ++++
MAINTAINERS | 9 +
drivers/pci/host/Kconfig | 11 +
drivers/pci/host/Makefile | 2 +
drivers/pci/host/pci-host-common.c | 194 ++++++++++++++
drivers/pci/host/pci-host-common.h | 56 ++++
drivers/pci/host/pci-host-generic.c | 190 +-------------
drivers/pci/host/pci-thunder-pem.c | 283 +++++++++++++++++++++
8 files changed, 602 insertions(+), 186 deletions(-)
create mode 100644 Documentation/devicetree/bindings/pci/pci-thunder-pem.txt
create mode 100644 drivers/pci/host/pci-host-common.c
create mode 100644 drivers/pci/host/pci-host-common.h
create mode 100644 drivers/pci/host/pci-thunder-pem.c
--
1.8.3.1
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers.
2016-01-08 23:59 [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe David Daney
@ 2016-01-08 23:59 ` David Daney
2016-01-12 17:59 ` Will Deacon
2016-01-08 23:59 ` [PATCH v3 2/2] pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors David Daney
2016-01-12 21:53 ` [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe Arnd Bergmann
2 siblings, 1 reply; 6+ messages in thread
From: David Daney @ 2016-01-08 23:59 UTC (permalink / raw)
To: linux-arm-kernel
From: David Daney <david.daney@cavium.com>
No change in functionality.
Move structure definitions into a separate header file. Move common
code to new file with Kconfig machinery to build it. Split probe
function in to two parts:
- a small driver specific probe function (gen_pci_probe)
- a common probe that can be used by other drivers
(pci_host_common_probe)
Signed-off-by: David Daney <david.daney@cavium.com>
---
MAINTAINERS | 1 +
drivers/pci/host/Kconfig | 4 +
drivers/pci/host/Makefile | 1 +
drivers/pci/host/pci-host-common.c | 194 ++++++++++++++++++++++++++++++++++++
drivers/pci/host/pci-host-common.h | 56 +++++++++++
drivers/pci/host/pci-host-generic.c | 190 +----------------------------------
6 files changed, 260 insertions(+), 186 deletions(-)
create mode 100644 drivers/pci/host/pci-host-common.c
create mode 100644 drivers/pci/host/pci-host-common.h
diff --git a/MAINTAINERS b/MAINTAINERS
index 233f834..a3fc8b4 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8217,6 +8217,7 @@ L: linux-pci at vger.kernel.org
L: linux-arm-kernel at lists.infradead.org (moderated for non-subscribers)
S: Maintained
F: Documentation/devicetree/bindings/pci/host-generic-pci.txt
+F: drivers/pci/host/pci-host-common.c
F: drivers/pci/host/pci-host-generic.c
PCIE DRIVER FOR ST SPEAR13XX
diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig
index f131ba9..f171459 100644
--- a/drivers/pci/host/Kconfig
+++ b/drivers/pci/host/Kconfig
@@ -53,9 +53,13 @@ config PCI_RCAR_GEN2_PCIE
help
Say Y here if you want PCIe controller support on R-Car Gen2 SoCs.
+config PCI_HOST_COMMON
+ bool
+
config PCI_HOST_GENERIC
bool "Generic PCI host controller"
depends on (ARM || ARM64) && OF
+ select PCI_HOST_COMMON
help
Say Y here if you want to support a simple generic PCI host
controller, such as the one emulated by kvmtool.
diff --git a/drivers/pci/host/Makefile b/drivers/pci/host/Makefile
index 9d4d3c6..ba2c51f 100644
--- a/drivers/pci/host/Makefile
+++ b/drivers/pci/host/Makefile
@@ -6,6 +6,7 @@ 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_RCAR_GEN2_PCIE) += pcie-rcar.o
+obj-$(CONFIG_PCI_HOST_COMMON) += pci-host-common.o
obj-$(CONFIG_PCI_HOST_GENERIC) += pci-host-generic.o
obj-$(CONFIG_PCIE_SPEAR13XX) += pcie-spear13xx.o
obj-$(CONFIG_PCI_KEYSTONE) += pci-keystone-dw.o pci-keystone.o
diff --git a/drivers/pci/host/pci-host-common.c b/drivers/pci/host/pci-host-common.c
new file mode 100644
index 0000000..e9f850f
--- /dev/null
+++ b/drivers/pci/host/pci-host-common.c
@@ -0,0 +1,194 @@
+/*
+ * 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@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>
+
+#include "pci-host-common.h"
+
+static void gen_pci_release_of_pci_ranges(struct gen_pci *pci)
+{
+ pci_free_resource_list(&pci->resources);
+}
+
+static int gen_pci_parse_request_of_pci_ranges(struct gen_pci *pci)
+{
+ int err, res_valid = 0;
+ struct device *dev = pci->host.dev.parent;
+ struct device_node *np = dev->of_node;
+ resource_size_t iobase;
+ struct resource_entry *win;
+
+ err = of_pci_get_host_bridge_resources(np, 0, 0xff, &pci->resources,
+ &iobase);
+ if (err)
+ return err;
+
+ resource_list_for_each_entry(win, &pci->resources) {
+ struct resource *parent, *res = win->res;
+
+ switch (resource_type(res)) {
+ case IORESOURCE_IO:
+ parent = &ioport_resource;
+ err = pci_remap_iospace(res, iobase);
+ if (err) {
+ dev_warn(dev, "error %d: failed to map resource %pR\n",
+ err, res);
+ continue;
+ }
+ break;
+ case IORESOURCE_MEM:
+ parent = &iomem_resource;
+ res_valid |= !(res->flags & IORESOURCE_PREFETCH);
+ break;
+ case IORESOURCE_BUS:
+ pci->cfg.bus_range = res;
+ default:
+ continue;
+ }
+
+ err = devm_request_resource(dev, parent, res);
+ if (err)
+ goto out_release_res;
+ }
+
+ if (!res_valid) {
+ dev_err(dev, "non-prefetchable memory resource required\n");
+ err = -EINVAL;
+ goto out_release_res;
+ }
+
+ return 0;
+
+out_release_res:
+ gen_pci_release_of_pci_ranges(pci);
+ return err;
+}
+
+static int gen_pci_parse_map_cfg_windows(struct gen_pci *pci)
+{
+ int err;
+ u8 bus_max;
+ resource_size_t busn;
+ struct resource *bus_range;
+ struct device *dev = pci->host.dev.parent;
+ struct device_node *np = dev->of_node;
+ u32 sz = 1 << pci->cfg.ops->bus_shift;
+
+ err = of_address_to_resource(np, 0, &pci->cfg.res);
+ if (err) {
+ dev_err(dev, "missing \"reg\" property\n");
+ return err;
+ }
+
+ /* 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);
+
+ pci->cfg.win = devm_kcalloc(dev, resource_size(pci->cfg.bus_range),
+ sizeof(*pci->cfg.win), GFP_KERNEL);
+ if (!pci->cfg.win)
+ return -ENOMEM;
+
+ /* Map our Configuration Space windows */
+ if (!devm_request_mem_region(dev, pci->cfg.res.start,
+ resource_size(&pci->cfg.res),
+ "Configuration Space"))
+ return -ENOMEM;
+
+ bus_range = pci->cfg.bus_range;
+ for (busn = bus_range->start; busn <= bus_range->end; ++busn) {
+ u32 idx = busn - bus_range->start;
+
+ pci->cfg.win[idx] = devm_ioremap(dev,
+ pci->cfg.res.start + idx * sz,
+ sz);
+ if (!pci->cfg.win[idx])
+ return -ENOMEM;
+ }
+
+ return 0;
+}
+
+int pci_host_common_probe(struct platform_device *pdev,
+ struct gen_pci *pci)
+{
+ int err;
+ const char *type;
+ struct device *dev = &pdev->dev;
+ struct device_node *np = dev->of_node;
+ struct pci_bus *bus, *child;
+
+ type = of_get_property(np, "device_type", NULL);
+ if (!type || strcmp(type, "pci")) {
+ dev_err(dev, "invalid \"device_type\" %s\n", type);
+ return -EINVAL;
+ }
+
+ of_pci_check_probe_only();
+
+ pci->host.dev.parent = dev;
+ INIT_LIST_HEAD(&pci->host.windows);
+ INIT_LIST_HEAD(&pci->resources);
+
+ /* Parse our PCI ranges and request their resources */
+ err = gen_pci_parse_request_of_pci_ranges(pci);
+ if (err)
+ return err;
+
+ /* Parse and map our Configuration Space windows */
+ err = gen_pci_parse_map_cfg_windows(pci);
+ if (err) {
+ gen_pci_release_of_pci_ranges(pci);
+ return err;
+ }
+
+ /* Do not reassign resources if probe only */
+ if (!pci_has_flag(PCI_PROBE_ONLY))
+ pci_add_flags(PCI_REASSIGN_ALL_RSRC | PCI_REASSIGN_ALL_BUS);
+
+
+ bus = pci_scan_root_bus(dev, pci->cfg.bus_range->start,
+ &pci->cfg.ops->ops, pci, &pci->resources);
+ if (!bus) {
+ dev_err(dev, "Scanning rootbus failed");
+ return -ENODEV;
+ }
+
+ pci_fixup_irqs(pci_common_swizzle, of_irq_parse_and_map_pci);
+
+ if (!pci_has_flag(PCI_PROBE_ONLY)) {
+ pci_bus_size_bridges(bus);
+ pci_bus_assign_resources(bus);
+
+ list_for_each_entry(child, &bus->children, node)
+ pcie_bus_configure_settings(child);
+ }
+
+ pci_bus_add_devices(bus);
+ return 0;
+}
+
+MODULE_DESCRIPTION("Generic PCI host driver common code");
+MODULE_AUTHOR("Will Deacon <will.deacon@arm.com>");
+MODULE_LICENSE("GPL v2");
diff --git a/drivers/pci/host/pci-host-common.h b/drivers/pci/host/pci-host-common.h
new file mode 100644
index 0000000..9423808
--- /dev/null
+++ b/drivers/pci/host/pci-host-common.h
@@ -0,0 +1,56 @@
+/*
+ * 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@arm.com>
+ */
+
+#ifndef _PCI_HOST_COMMON_H
+#define _PCI_HOST_COMMON_H
+
+#include <linux/kernel.h>
+#include <linux/platform_device.h>
+
+struct gen_pci_cfg_bus_ops {
+ u32 bus_shift;
+ struct pci_ops ops;
+};
+
+struct gen_pci_cfg_windows {
+ struct resource res;
+ struct resource *bus_range;
+ void __iomem **win;
+
+ struct gen_pci_cfg_bus_ops *ops;
+};
+
+/*
+ * ARM pcibios functions expect the ARM struct pci_sys_data as the PCI
+ * sysdata. Add pci_sys_data as the first element in struct gen_pci so
+ * that when we use a gen_pci pointer as sysdata, it is also a pointer to
+ * a struct pci_sys_data.
+ */
+struct gen_pci {
+#ifdef CONFIG_ARM
+ struct pci_sys_data sys;
+#endif
+ struct pci_host_bridge host;
+ struct gen_pci_cfg_windows cfg;
+ struct list_head resources;
+};
+
+int pci_host_common_probe(struct platform_device *pdev,
+ struct gen_pci *pci);
+
+#endif /* _PCI_HOST_COMMON_H */
diff --git a/drivers/pci/host/pci-host-generic.c b/drivers/pci/host/pci-host-generic.c
index 5434c90..e8aa78f 100644
--- a/drivers/pci/host/pci-host-generic.c
+++ b/drivers/pci/host/pci-host-generic.c
@@ -25,33 +25,7 @@
#include <linux/of_pci.h>
#include <linux/platform_device.h>
-struct gen_pci_cfg_bus_ops {
- u32 bus_shift;
- struct pci_ops ops;
-};
-
-struct gen_pci_cfg_windows {
- struct resource res;
- struct resource *bus_range;
- void __iomem **win;
-
- struct gen_pci_cfg_bus_ops *ops;
-};
-
-/*
- * ARM pcibios functions expect the ARM struct pci_sys_data as the PCI
- * sysdata. Add pci_sys_data as the first element in struct gen_pci so
- * that when we use a gen_pci pointer as sysdata, it is also a pointer to
- * a struct pci_sys_data.
- */
-struct gen_pci {
-#ifdef CONFIG_ARM
- struct pci_sys_data sys;
-#endif
- struct pci_host_bridge host;
- struct gen_pci_cfg_windows cfg;
- struct list_head resources;
-};
+#include "pci-host-common.h"
static void __iomem *gen_pci_map_cfg_bus_cam(struct pci_bus *bus,
unsigned int devfn,
@@ -102,175 +76,19 @@ static const struct of_device_id gen_pci_of_match[] = {
};
MODULE_DEVICE_TABLE(of, gen_pci_of_match);
-static void gen_pci_release_of_pci_ranges(struct gen_pci *pci)
-{
- pci_free_resource_list(&pci->resources);
-}
-
-static int gen_pci_parse_request_of_pci_ranges(struct gen_pci *pci)
-{
- int err, res_valid = 0;
- struct device *dev = pci->host.dev.parent;
- struct device_node *np = dev->of_node;
- resource_size_t iobase;
- struct resource_entry *win;
-
- err = of_pci_get_host_bridge_resources(np, 0, 0xff, &pci->resources,
- &iobase);
- if (err)
- return err;
-
- resource_list_for_each_entry(win, &pci->resources) {
- struct resource *parent, *res = win->res;
-
- switch (resource_type(res)) {
- case IORESOURCE_IO:
- parent = &ioport_resource;
- err = pci_remap_iospace(res, iobase);
- if (err) {
- dev_warn(dev, "error %d: failed to map resource %pR\n",
- err, res);
- continue;
- }
- break;
- case IORESOURCE_MEM:
- parent = &iomem_resource;
- res_valid |= !(res->flags & IORESOURCE_PREFETCH);
- break;
- case IORESOURCE_BUS:
- pci->cfg.bus_range = res;
- default:
- continue;
- }
-
- err = devm_request_resource(dev, parent, res);
- if (err)
- goto out_release_res;
- }
-
- if (!res_valid) {
- dev_err(dev, "non-prefetchable memory resource required\n");
- err = -EINVAL;
- goto out_release_res;
- }
-
- return 0;
-
-out_release_res:
- gen_pci_release_of_pci_ranges(pci);
- return err;
-}
-
-static int gen_pci_parse_map_cfg_windows(struct gen_pci *pci)
-{
- int err;
- u8 bus_max;
- resource_size_t busn;
- struct resource *bus_range;
- struct device *dev = pci->host.dev.parent;
- struct device_node *np = dev->of_node;
- u32 sz = 1 << pci->cfg.ops->bus_shift;
-
- err = of_address_to_resource(np, 0, &pci->cfg.res);
- if (err) {
- dev_err(dev, "missing \"reg\" property\n");
- return err;
- }
-
- /* 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);
-
- pci->cfg.win = devm_kcalloc(dev, resource_size(pci->cfg.bus_range),
- sizeof(*pci->cfg.win), GFP_KERNEL);
- if (!pci->cfg.win)
- return -ENOMEM;
-
- /* Map our Configuration Space windows */
- if (!devm_request_mem_region(dev, pci->cfg.res.start,
- resource_size(&pci->cfg.res),
- "Configuration Space"))
- return -ENOMEM;
-
- bus_range = pci->cfg.bus_range;
- for (busn = bus_range->start; busn <= bus_range->end; ++busn) {
- u32 idx = busn - bus_range->start;
-
- pci->cfg.win[idx] = devm_ioremap(dev,
- pci->cfg.res.start + idx * sz,
- sz);
- if (!pci->cfg.win[idx])
- return -ENOMEM;
- }
-
- return 0;
-}
-
static int gen_pci_probe(struct platform_device *pdev)
{
- int err;
- const char *type;
- const struct of_device_id *of_id;
struct device *dev = &pdev->dev;
- struct device_node *np = dev->of_node;
+ const struct of_device_id *of_id;
struct gen_pci *pci = devm_kzalloc(dev, sizeof(*pci), GFP_KERNEL);
- struct pci_bus *bus, *child;
if (!pci)
return -ENOMEM;
- type = of_get_property(np, "device_type", NULL);
- if (!type || strcmp(type, "pci")) {
- dev_err(dev, "invalid \"device_type\" %s\n", type);
- return -EINVAL;
- }
-
- of_pci_check_probe_only();
-
- of_id = of_match_node(gen_pci_of_match, np);
+ of_id = of_match_node(gen_pci_of_match, dev->of_node);
pci->cfg.ops = (struct gen_pci_cfg_bus_ops *)of_id->data;
- pci->host.dev.parent = dev;
- INIT_LIST_HEAD(&pci->host.windows);
- INIT_LIST_HEAD(&pci->resources);
-
- /* Parse our PCI ranges and request their resources */
- err = gen_pci_parse_request_of_pci_ranges(pci);
- if (err)
- return err;
-
- /* Parse and map our Configuration Space windows */
- err = gen_pci_parse_map_cfg_windows(pci);
- if (err) {
- gen_pci_release_of_pci_ranges(pci);
- return err;
- }
-
- /* Do not reassign resources if probe only */
- if (!pci_has_flag(PCI_PROBE_ONLY))
- pci_add_flags(PCI_REASSIGN_ALL_RSRC | PCI_REASSIGN_ALL_BUS);
-
-
- bus = pci_scan_root_bus(dev, pci->cfg.bus_range->start,
- &pci->cfg.ops->ops, pci, &pci->resources);
- if (!bus) {
- dev_err(dev, "Scanning rootbus failed");
- return -ENODEV;
- }
-
- pci_fixup_irqs(pci_common_swizzle, of_irq_parse_and_map_pci);
-
- if (!pci_has_flag(PCI_PROBE_ONLY)) {
- pci_bus_size_bridges(bus);
- pci_bus_assign_resources(bus);
-
- list_for_each_entry(child, &bus->children, node)
- pcie_bus_configure_settings(child);
- }
- pci_bus_add_devices(bus);
- return 0;
+ return pci_host_common_probe(pdev, pci);
}
static struct platform_driver gen_pci_driver = {
--
1.8.3.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH v3 2/2] pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors.
2016-01-08 23:59 [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe David Daney
2016-01-08 23:59 ` [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers David Daney
@ 2016-01-08 23:59 ` David Daney
2016-01-09 22:53 ` Rob Herring
2016-01-12 21:53 ` [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe Arnd Bergmann
2 siblings, 1 reply; 6+ messages in thread
From: David Daney @ 2016-01-08 23:59 UTC (permalink / raw)
To: linux-arm-kernel
From: David Daney <david.daney@cavium.com>
Some Cavium ThunderX processors require quirky access methods for the
config space of the PCIe bridge. Add a driver to provide these config
space accessor functions. The pci-host-common code is used to
configure the PCI machinery.
Signed-off-by: David Daney <david.daney@cavium.com>
---
.../devicetree/bindings/pci/pci-thunder-pem.txt | 43 ++++
MAINTAINERS | 8 +
drivers/pci/host/Kconfig | 7 +
drivers/pci/host/Makefile | 1 +
drivers/pci/host/pci-thunder-pem.c | 283 +++++++++++++++++++++
5 files changed, 342 insertions(+)
create mode 100644 Documentation/devicetree/bindings/pci/pci-thunder-pem.txt
create mode 100644 drivers/pci/host/pci-thunder-pem.c
diff --git a/Documentation/devicetree/bindings/pci/pci-thunder-pem.txt b/Documentation/devicetree/bindings/pci/pci-thunder-pem.txt
new file mode 100644
index 0000000..f131fae
--- /dev/null
+++ b/Documentation/devicetree/bindings/pci/pci-thunder-pem.txt
@@ -0,0 +1,43 @@
+* ThunderX PEM PCIe host controller
+
+Firmware-initialized PCI host controller found on some Cavium
+ThunderX processors.
+
+The properties and their meanings are identical to those described in
+host-generic-pci.txt except as listed below.
+
+Properties of the host controller node that differ from
+host-generic-pci.txt:
+
+- compatible : Must be "cavium,pci-host-thunder-pem"
+
+- reg : Two entries: First the configuration space for down
+ stream devices base address and size, as accessed
+ from the parent bus. Second, the register bank of
+ the PEM device PCIe bridge.
+
+Example:
+
+ pci at 87e0,c2000000 {
+ compatible = "cavium,pci-host-thunder-pem";
+ device_type = "pci";
+ msi-parent = <&its>;
+ msi-map = <0 &its 0x10000 0x10000>;
+ bus-range = <0x8f 0xc7>;
+ #size-cells = <2>;
+ #address-cells = <3>;
+
+ reg = <0x8880 0x8f000000 0x0 0x39000000>, /* Configuration space */
+ <0x87e0 0xc2000000 0x0 0x00010000>; /* PEM space */
+ ranges = <0x01000000 0x00 0x00020000 0x88b0 0x00020000 0x00 0x00010000>, /* I/O */
+ <0x03000000 0x00 0x10000000 0x8890 0x10000000 0x0f 0xf0000000>, /* mem64 */
+ <0x43000000 0x10 0x00000000 0x88a0 0x00000000 0x10 0x00000000>, /* mem64-pref */
+ <0x03000000 0x87e0 0xc2f00000 0x87e0 0xc2000000 0x00 0x00100000>; /* mem64 PEM BAR4 */
+
+ #interrupt-cells = <1>;
+ interrupt-map-mask = <0 0 0 7>;
+ interrupt-map = <0 0 0 1 &gic0 0 0 0 24 4>, /* INTA */
+ <0 0 0 2 &gic0 0 0 0 25 4>, /* INTB */
+ <0 0 0 3 &gic0 0 0 0 26 4>, /* INTC */
+ <0 0 0 4 &gic0 0 0 0 27 4>; /* INTD */
+ };
diff --git a/MAINTAINERS b/MAINTAINERS
index a3fc8b4..24a04cc 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -8249,6 +8249,14 @@ S: Maintained
F: Documentation/devicetree/bindings/pci/hisilicon-pcie.txt
F: drivers/pci/host/pcie-hisi.c
+PCIE DRIVER FOR CAVIUM THUNDERX
+M: David Daney <david.daney@cavium.com>
+L: linux-pci at vger.kernel.org
+L: linux-arm-kernel at lists.infradead.org (moderated for non-subscribers)
+S: Supported
+F: Documentation/devicetree/bindings/pci/pci-thunder-pem.txt
+F: drivers/pci/host/pci-thunder-pem.c
+
PCMCIA SUBSYSTEM
P: Linux PCMCIA Team
L: linux-pcmcia at lists.infradead.org
diff --git a/drivers/pci/host/Kconfig b/drivers/pci/host/Kconfig
index f171459..ad32f251 100644
--- a/drivers/pci/host/Kconfig
+++ b/drivers/pci/host/Kconfig
@@ -176,4 +176,11 @@ config PCI_HISI
help
Say Y here if you want PCIe controller support on HiSilicon HIP05 SoC
+config PCI_HOST_THUNDER_PEM
+ bool "Cavium Thunder PCIe controller to off-chip devices"
+ depends on OF && ARM64
+ select PCI_HOST_COMMON
+ help
+ Say Y here if you want PCIe support for CN88XX Cavium Thunder SoCs.
+
endmenu
diff --git a/drivers/pci/host/Makefile b/drivers/pci/host/Makefile
index ba2c51f..a068d10 100644
--- a/drivers/pci/host/Makefile
+++ b/drivers/pci/host/Makefile
@@ -21,3 +21,4 @@ obj-$(CONFIG_PCIE_IPROC_BCMA) += pcie-iproc-bcma.o
obj-$(CONFIG_PCIE_ALTERA) += pcie-altera.o
obj-$(CONFIG_PCIE_ALTERA_MSI) += pcie-altera-msi.o
obj-$(CONFIG_PCI_HISI) += pcie-hisi.o
+obj-$(CONFIG_PCI_HOST_THUNDER_PEM) += pci-thunder-pem.o
diff --git a/drivers/pci/host/pci-thunder-pem.c b/drivers/pci/host/pci-thunder-pem.c
new file mode 100644
index 0000000..43fa6f5
--- /dev/null
+++ b/drivers/pci/host/pci-thunder-pem.c
@@ -0,0 +1,283 @@
+/*
+ * 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) 2015 Cavium, Inc.
+ *
+ */
+
+#include <linux/kernel.h>
+#include <linux/module.h>
+#include <linux/of_address.h>
+#include <linux/of_pci.h>
+#include <linux/platform_device.h>
+
+#include "pci-host-common.h"
+
+#define PEM_CFG_WR 0x28
+#define PEM_CFG_RD 0x30
+
+struct thunder_pem_pci {
+ struct gen_pci gen_pci;
+ u32 ea_entry[3];
+ void __iomem *pem_reg_base;
+};
+
+static int thunder_pem_config_read(struct pci_bus *bus, unsigned int devfn,
+ int where, int size, u32 *val)
+{
+ int r;
+ struct thunder_pem_pci *pem_pci;
+ struct gen_pci *pci = bus->sysdata;
+
+ pem_pci = container_of(pci, struct thunder_pem_pci, gen_pci);
+
+ /*
+ * The first device on the bus in the PEM PCIe bridge.
+ * Special case its config access.
+ */
+ if (bus->number == pci->cfg.bus_range->start) {
+ u64 read_val;
+
+ if (devfn != 0 || where >= 2048) {
+ *val = ~0;
+ return PCIBIOS_DEVICE_NOT_FOUND;
+ }
+
+ /*
+ * 32-bit accesses only. Write the address to the low
+ * order bits of PEM_CFG_RD, then trigger the read by
+ * reading back. The config data lands in the upper
+ * 32-bits of PEM_CFG_RD.
+ */
+ read_val = where & ~3ull;
+ writeq(read_val, pem_pci->pem_reg_base + PEM_CFG_RD);
+ read_val = readq(pem_pci->pem_reg_base + PEM_CFG_RD);
+ read_val >>= 32;
+
+ /*
+ * The config space contains some garbage, fix it up.
+ * Also synthesize an EA capability for the BAR used
+ * by MSI-X.
+ */
+ switch (where & ~3u) {
+ case 0x40:
+ read_val &= 0xffff00ff;
+ read_val |= 0x00007000; /* Skip MSI CAP */
+ break;
+ case 0x70: /* Express Cap */
+ /* PME interrupt on vector 2*/
+ read_val |= (2u << 25);
+ break;
+ case 0xb0: /* MSI-X Cap */
+ /* TableSize=4, Next Cap is EA */
+ read_val &= 0xc00000ff;
+ read_val |= 0x0003bc00;
+ break;
+ case 0xb4:
+ /* Table offset=0, BIR=0 */
+ read_val = 0x00000000;
+ break;
+ case 0xb8:
+ /* BPA offset=0xf0000, BIR=0 */
+ read_val = 0x000f0000;
+ break;
+ case 0xbc:
+ /* EA, 1 entry, no next Cap */
+ read_val = 0x00010014;
+ break;
+ case 0xc0:
+ /* DW2 for type-1 */
+ read_val = 0x00000000;
+ break;
+ case 0xc4:
+ /* Entry BEI=0, PP=0x00, SP=0xff, ES=3 */
+ read_val = 0x80ff0003;
+ break;
+ case 0xc8:
+ read_val = pem_pci->ea_entry[0];
+ break;
+ case 0xcc:
+ read_val = pem_pci->ea_entry[1];
+ break;
+ case 0xd0:
+ read_val = pem_pci->ea_entry[2];
+ break;
+ default:
+ break;
+ }
+ read_val >>= (8 * (where & 3));
+ switch (size) {
+ case 1:
+ read_val &= 0xff;
+ break;
+ case 2:
+ read_val &= 0xffff;
+ break;
+ default:
+ break;
+ }
+ *val = read_val;
+ return PCIBIOS_SUCCESSFUL;
+ }
+ if (bus->number < pci->cfg.bus_range->start ||
+ bus->number > pci->cfg.bus_range->end)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+
+ r = pci_generic_config_read(bus, devfn, where, size, val);
+ return r;
+}
+
+static int thunder_pem_config_write(struct pci_bus *bus, unsigned int devfn,
+ int where, int size, u32 val)
+{
+ struct gen_pci *pci = bus->sysdata;
+ struct thunder_pem_pci *pem_pci;
+
+ pem_pci = container_of(pci, struct thunder_pem_pci, gen_pci);
+
+ /*
+ * The first device on the bus in the PEM PCIe bridge.
+ * Special case its config access.
+ */
+ if (bus->number == pci->cfg.bus_range->start) {
+ u64 write_val, read_val;
+
+ if (devfn != 0 || where >= 2048)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+
+ /*
+ * 32-bit accesses only. If the write is for a size
+ * smaller than 32-bits, we must first read the 32-bit
+ * value and merge in the desired bits and then write
+ * the whole 32-bits back out.
+ */
+ switch (size) {
+ case 1:
+ read_val = where & ~3ull;
+ writeq(read_val, pem_pci->pem_reg_base + PEM_CFG_RD);
+ read_val = readq(pem_pci->pem_reg_base + PEM_CFG_RD);
+ read_val >>= 32;
+ read_val &= ~(0xff << (8 * (where & 3)));
+ val = (val & 0xff) << (8 * (where & 3));
+ val |= (u32)read_val;
+ break;
+ case 2:
+ read_val = where & ~3ull;
+ writeq(read_val, pem_pci->pem_reg_base + PEM_CFG_RD);
+ read_val = readq(pem_pci->pem_reg_base + PEM_CFG_RD);
+ read_val >>= 32;
+ read_val &= ~(0xffff << (8 * (where & 3)));
+ val = (val & 0xffff) << (8 * (where & 3));
+ val |= (u32)read_val;
+ break;
+ default:
+ break;
+
+ }
+ /*
+ * Low order bits are the config address, the high
+ * order 32 bits are the data to be written.
+ */
+ write_val = where & ~3ull;
+ write_val |= (((u64)val) << 32);
+ writeq(write_val, pem_pci->pem_reg_base + PEM_CFG_WR);
+ return PCIBIOS_SUCCESSFUL;
+ }
+ if (bus->number < pci->cfg.bus_range->start ||
+ bus->number > pci->cfg.bus_range->end)
+ return PCIBIOS_DEVICE_NOT_FOUND;
+ return pci_generic_config_write(bus, devfn, where, size, val);
+}
+
+static void __iomem *map_cfg_bus_thunder_pem(struct pci_bus *bus,
+ unsigned int devfn,
+ int where)
+{
+ struct gen_pci *pci = bus->sysdata;
+ resource_size_t idx = bus->number - pci->cfg.bus_range->start;
+
+ return pci->cfg.win[idx] + ((devfn << 16) | where);
+}
+
+static struct gen_pci_cfg_bus_ops thunder_pem_bus_ops = {
+ .bus_shift = 24,
+ .ops = {
+ .map_bus = map_cfg_bus_thunder_pem,
+ .read = thunder_pem_config_read,
+ .write = thunder_pem_config_write,
+ }
+};
+
+static const struct of_device_id thunder_pem_of_match[] = {
+ { .compatible = "cavium,pci-host-thunder-pem",
+ .data = &thunder_pem_bus_ops },
+
+ { },
+};
+MODULE_DEVICE_TABLE(of, thunder_pem_of_match);
+
+static int thunder_pem_probe(struct platform_device *pdev)
+{
+ struct device *dev = &pdev->dev;
+ const struct of_device_id *of_id;
+ resource_size_t bar4_start;
+ struct resource *res_pem;
+ struct thunder_pem_pci *pem_pci;
+
+ pem_pci = devm_kzalloc(dev, sizeof(*pem_pci), GFP_KERNEL);
+ if (!pem_pci)
+ return -ENOMEM;
+
+ of_id = of_match_node(thunder_pem_of_match, dev->of_node);
+ pem_pci->gen_pci.cfg.ops = (struct gen_pci_cfg_bus_ops *)of_id->data;
+
+ /*
+ * The second register range is the PEM bridge to the PCIe
+ * bus. It has a different config access method than those
+ * devices behind the bridge.
+ */
+ res_pem = platform_get_resource(pdev, IORESOURCE_MEM, 1);
+ if (!res_pem) {
+ dev_err(dev, "missing \"reg[1]\"property\n");
+ return -EINVAL;
+ }
+
+ pem_pci->pem_reg_base = devm_ioremap(dev, res_pem->start, 0x10000);
+ if (!pem_pci->pem_reg_base)
+ return -ENOMEM;
+
+ /*
+ * The MSI-X BAR for the PEM and AER interrupts is located at
+ * a fixed offset from the PEM register base. Generate a
+ * fragment of the synthesized Enhanced Allocation capability
+ * structure here for the BAR.
+ */
+ bar4_start = res_pem->start + 0xf00000;
+ pem_pci->ea_entry[0] = (u32)bar4_start | 2;
+ pem_pci->ea_entry[1] = (u32)(res_pem->end - bar4_start) & ~3u;
+ pem_pci->ea_entry[2] = (u32)(bar4_start >> 32);
+
+ return pci_host_common_probe(pdev, &pem_pci->gen_pci);
+}
+
+static struct platform_driver thunder_pem_driver = {
+ .driver = {
+ .name = KBUILD_MODNAME,
+ .of_match_table = thunder_pem_of_match,
+ },
+ .probe = thunder_pem_probe,
+};
+module_platform_driver(thunder_pem_driver);
+
+MODULE_DESCRIPTION("Thunder PEM PCIe host driver");
+MODULE_LICENSE("GPL v2");
--
1.8.3.1
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH v3 2/2] pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors.
2016-01-08 23:59 ` [PATCH v3 2/2] pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors David Daney
@ 2016-01-09 22:53 ` Rob Herring
0 siblings, 0 replies; 6+ messages in thread
From: Rob Herring @ 2016-01-09 22:53 UTC (permalink / raw)
To: linux-arm-kernel
On Fri, Jan 08, 2016 at 03:59:13PM -0800, David Daney wrote:
> From: David Daney <david.daney@cavium.com>
>
> Some Cavium ThunderX processors require quirky access methods for the
> config space of the PCIe bridge. Add a driver to provide these config
> space accessor functions. The pci-host-common code is used to
> configure the PCI machinery.
>
> Signed-off-by: David Daney <david.daney@cavium.com>
> ---
> .../devicetree/bindings/pci/pci-thunder-pem.txt | 43 ++++
Acked-by: Rob Herring <robh@kernel.org>
> MAINTAINERS | 8 +
> drivers/pci/host/Kconfig | 7 +
> drivers/pci/host/Makefile | 1 +
> drivers/pci/host/pci-thunder-pem.c | 283 +++++++++++++++++++++
> 5 files changed, 342 insertions(+)
> create mode 100644 Documentation/devicetree/bindings/pci/pci-thunder-pem.txt
> create mode 100644 drivers/pci/host/pci-thunder-pem.c
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers.
2016-01-08 23:59 ` [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers David Daney
@ 2016-01-12 17:59 ` Will Deacon
0 siblings, 0 replies; 6+ messages in thread
From: Will Deacon @ 2016-01-12 17:59 UTC (permalink / raw)
To: linux-arm-kernel
On Fri, Jan 08, 2016 at 03:59:12PM -0800, David Daney wrote:
> From: David Daney <david.daney@cavium.com>
>
> No change in functionality.
>
> Move structure definitions into a separate header file. Move common
> code to new file with Kconfig machinery to build it. Split probe
> function in to two parts:
>
> - a small driver specific probe function (gen_pci_probe)
>
> - a common probe that can be used by other drivers
> (pci_host_common_probe)
>
> Signed-off-by: David Daney <david.daney@cavium.com>
> ---
> MAINTAINERS | 1 +
> drivers/pci/host/Kconfig | 4 +
> drivers/pci/host/Makefile | 1 +
> drivers/pci/host/pci-host-common.c | 194 ++++++++++++++++++++++++++++++++++++
> drivers/pci/host/pci-host-common.h | 56 +++++++++++
> drivers/pci/host/pci-host-generic.c | 190 +----------------------------------
> 6 files changed, 260 insertions(+), 186 deletions(-)
> create mode 100644 drivers/pci/host/pci-host-common.c
> create mode 100644 drivers/pci/host/pci-host-common.h
Looks fine to me:
Acked-by: Will Deacon <will.deacon@arm.com>
Will
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe
2016-01-08 23:59 [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe David Daney
2016-01-08 23:59 ` [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers David Daney
2016-01-08 23:59 ` [PATCH v3 2/2] pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors David Daney
@ 2016-01-12 21:53 ` Arnd Bergmann
2 siblings, 0 replies; 6+ messages in thread
From: Arnd Bergmann @ 2016-01-12 21:53 UTC (permalink / raw)
To: linux-arm-kernel
On Friday 08 January 2016 15:59:11 David Daney wrote:
> From: David Daney <david.daney@cavium.com>
>
> Some Cavium ThunderX processors require quirky access methods for the
> config space of the PCIe bridge.
>
> There are two patches:
>
> 1) Refactor code in pci-host-generic so that it can more easily be
> used by other drivers. This splits the driver for CAM and ECAM
> access methods to a separate file from the common host driver code.
>
> 2) Add the ThunderX PCIe driver, which leverages the code in
> pci-host-generic
>
> Changes from v2: Improve device tree binding example as noted by Rob
> Herring. Rename pcie-thunder-pem.* to pci-thunder-pem.* for better
> consistency. Update MAINTAINERS to reflect the changes.
>
> Changes from v1: Split CAM and ECAM code from common driver code as
> suggested by Arnd Bergmann. Fix spelling errors in
> pcie-thunder-pem.txt
>
Sorry for the late reply (I didn't comment on v2 either). This version
looks fine to me. There are probably cleanups we can do on the generic
code later on, but splitting it up the way you do here fully addresses
my earlier concerns. Thanks for your patience.
Acked-by: Arnd Bergmann <arnd@arndb.de>
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2016-01-12 21:53 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2016-01-08 23:59 [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe David Daney
2016-01-08 23:59 ` [PATCH v3 1/2] PCI: generic: Refactor code to enable reuse by other drivers David Daney
2016-01-12 17:59 ` Will Deacon
2016-01-08 23:59 ` [PATCH v3 2/2] pci, pci-thunder-pem: Add PCIe host driver for ThunderX processors David Daney
2016-01-09 22:53 ` Rob Herring
2016-01-12 21:53 ` [PATCH v3 0/2] pci: Add host controller driver for Cavium ThunderX PCIe Arnd Bergmann
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).