Hi All,
The Programmable Real-Time Unit and Industrial Communication Subsystem
(PRU-ICSS or simply PRUSS) on various TI SoCs consists of dual 32-bit
RISC cores (Programmable Real-Time Units, or PRUs) for program execution.
The K3 AM65x amd J721E SoCs have the next generation of the PRU-ICSS IP,
commonly called ICSSG. The ICSSG IP on AM65x SoCs has two PRU cores,
two auxiliary custom PRU cores called Real Time Units (RTUs). The K3
AM65x SR2.0 and J721E SoCs have a revised version of the ICSSG IP, and
include two additional custom auxiliary PRU cores called Transmit PRUs
(Tx_PRUs).
This series contains the PRUSS remoteproc driver together with relevant
dt-binding. This is the 3rd foundation component for PRUSS subsystem, the
previous two were already merged and can be found under:
1) drivers/soc/ti/pruss.c
Documentation/devicetree/bindings/soc/ti/ti,pruss.yaml
2) drivers/irqchip/irq-pruss-intc.c
Documentation/devicetree/bindings/interrupt-controller/ti,pruss-intc.yaml
Best regards,
Grzegorz
Grzegorz Jaszczyk (1):
remoteproc/pru: Add support for PRU specific interrupt configuration
Suman Anna (5):
dt-bindings: remoteproc: Add binding doc for PRU cores in the PRU-ICSS
remoteproc/pru: Add a PRU remoteproc driver
remoteproc/pru: Add pru-specific debugfs support
remoteproc/pru: Add support for various PRU cores on K3 AM65x SoCs
remoteproc/pru: Add support for various PRU cores on K3 J721E SoCs
.../bindings/remoteproc/ti,pru-rproc.yaml | 214 +++++
drivers/remoteproc/Kconfig | 12 +
drivers/remoteproc/Makefile | 1 +
drivers/remoteproc/pru_rproc.c | 880 ++++++++++++++++++
drivers/remoteproc/pru_rproc.h | 46 +
5 files changed, 1153 insertions(+)
create mode 100644 Documentation/devicetree/bindings/remoteproc/ti,pru-rproc.yaml
create mode 100644 drivers/remoteproc/pru_rproc.c
create mode 100644 drivers/remoteproc/pru_rproc.h
--
2.29.0
From: Suman Anna <[email protected]>
The Programmable Real-Time Unit Subsystem (PRUSS) consists of
dual 32-bit RISC cores (Programmable Real-Time Units, or PRUs)
for program execution. This patch adds a remoteproc platform
driver for managing the individual PRU RISC cores life cycle.
The PRUs do not have a unified address space (have an Instruction
RAM and a primary Data RAM at both 0x0). The PRU remoteproc driver
therefore uses a custom remoteproc core ELF loader ops. The added
.da_to_va ops is only used to provide translations for the PRU
Data RAMs. This remoteproc driver does not have support for error
recovery and system suspend/resume features. Different compatibles
are used to allow providing scalability for instance-specific device
data if needed. The driver uses a default firmware-name retrieved
from device-tree for each PRU core, and the firmwares are expected
to be present in the standard Linux firmware search paths. They can
also be adjusted by userspace if required through the sysfs interface
provided by the remoteproc core.
The PRU remoteproc driver uses a client-driven boot methodology: it
does _not_ support auto-boot so that the PRU load and boot is dictated
by the corresponding client drivers for achieving various usecases.
This allows flexibility for the client drivers or applications to set
a firmware name (if needed) based on their desired functionality and
boot the PRU. The sysfs bind and unbind attributes have also been
suppressed so that the PRU devices cannot be unbound and thereby
shutdown a PRU from underneath a PRU client driver.
The driver currently supports the AM335x, AM437x, AM57xx and 66AK2G
SoCs, and support for other TI SoCs will be added in subsequent
patches.
Co-developed-by: Andrew F. Davis <[email protected]>
Signed-off-by: Andrew F. Davis <[email protected]>
Signed-off-by: Suman Anna <[email protected]>
Co-developed-by: Grzegorz Jaszczyk <[email protected]>
Signed-off-by: Grzegorz Jaszczyk <[email protected]>
---
drivers/remoteproc/Kconfig | 12 +
drivers/remoteproc/Makefile | 1 +
drivers/remoteproc/pru_rproc.c | 428 +++++++++++++++++++++++++++++++++
3 files changed, 441 insertions(+)
create mode 100644 drivers/remoteproc/pru_rproc.c
diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig
index d99548fb5dde..3e3865a7cd78 100644
--- a/drivers/remoteproc/Kconfig
+++ b/drivers/remoteproc/Kconfig
@@ -125,6 +125,18 @@ config KEYSTONE_REMOTEPROC
It's safe to say N here if you're not interested in the Keystone
DSPs or just want to use a bare minimum kernel.
+config PRU_REMOTEPROC
+ tristate "TI PRU remoteproc support"
+ depends on TI_PRUSS
+ default TI_PRUSS
+ help
+ Support for TI PRU remote processors present within a PRU-ICSS
+ subsystem via the remote processor framework.
+
+ Say Y or M here to support the Programmable Realtime Unit (PRU)
+ processors on various TI SoCs. It's safe to say N here if you're
+ not interested in the PRU or if you are unsure.
+
config QCOM_PIL_INFO
tristate
diff --git a/drivers/remoteproc/Makefile b/drivers/remoteproc/Makefile
index da2ace4ec86c..bb26c9e4ef9c 100644
--- a/drivers/remoteproc/Makefile
+++ b/drivers/remoteproc/Makefile
@@ -18,6 +18,7 @@ obj-$(CONFIG_OMAP_REMOTEPROC) += omap_remoteproc.o
obj-$(CONFIG_WKUP_M3_RPROC) += wkup_m3_rproc.o
obj-$(CONFIG_DA8XX_REMOTEPROC) += da8xx_remoteproc.o
obj-$(CONFIG_KEYSTONE_REMOTEPROC) += keystone_remoteproc.o
+obj-$(CONFIG_PRU_REMOTEPROC) += pru_rproc.o
obj-$(CONFIG_QCOM_PIL_INFO) += qcom_pil_info.o
obj-$(CONFIG_QCOM_RPROC_COMMON) += qcom_common.o
obj-$(CONFIG_QCOM_Q6V5_COMMON) += qcom_q6v5.o
diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
new file mode 100644
index 000000000000..c94c8e965c21
--- /dev/null
+++ b/drivers/remoteproc/pru_rproc.c
@@ -0,0 +1,428 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * PRU-ICSS remoteproc driver for various TI SoCs
+ *
+ * Copyright (C) 2014-2020 Texas Instruments Incorporated - https://www.ti.com/
+ *
+ * Author(s):
+ * Suman Anna <[email protected]>
+ * Andrew F. Davis <[email protected]>
+ * Grzegorz Jaszczyk <[email protected]> for Texas Instruments
+ */
+
+#include <linux/bitops.h>
+#include <linux/module.h>
+#include <linux/of_device.h>
+#include <linux/pruss_driver.h>
+#include <linux/remoteproc.h>
+
+#include "remoteproc_internal.h"
+#include "remoteproc_elf_helpers.h"
+
+/* PRU_ICSS_PRU_CTRL registers */
+#define PRU_CTRL_CTRL 0x0000
+#define PRU_CTRL_STS 0x0004
+
+/* CTRL register bit-fields */
+#define CTRL_CTRL_SOFT_RST_N BIT(0)
+#define CTRL_CTRL_EN BIT(1)
+#define CTRL_CTRL_SLEEPING BIT(2)
+#define CTRL_CTRL_CTR_EN BIT(3)
+#define CTRL_CTRL_SINGLE_STEP BIT(8)
+#define CTRL_CTRL_RUNSTATE BIT(15)
+
+/* PRU Core IRAM address masks */
+#define PRU0_IRAM_ADDR_MASK 0x34000
+#define PRU1_IRAM_ADDR_MASK 0x38000
+
+/* PRU device addresses for various type of PRU RAMs */
+#define PRU_IRAM_DA 0 /* Instruction RAM */
+#define PRU_PDRAM_DA 0 /* Primary Data RAM */
+#define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
+#define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
+
+/**
+ * enum pru_iomem - PRU core memory/register range identifiers
+ *
+ * @PRU_IOMEM_IRAM: PRU Instruction RAM range
+ * @PRU_IOMEM_CTRL: PRU Control register range
+ * @PRU_IOMEM_DEBUG: PRU Debug register range
+ * @PRU_IOMEM_MAX: just keep this one at the end
+ */
+enum pru_iomem {
+ PRU_IOMEM_IRAM = 0,
+ PRU_IOMEM_CTRL,
+ PRU_IOMEM_DEBUG,
+ PRU_IOMEM_MAX,
+};
+
+/**
+ * struct pru_rproc - PRU remoteproc structure
+ * @id: id of the PRU core within the PRUSS
+ * @dev: PRU core device pointer
+ * @pruss: back-reference to parent PRUSS structure
+ * @rproc: remoteproc pointer for this PRU core
+ * @mem_regions: data for each of the PRU memory regions
+ * @fw_name: name of firmware image used during loading
+ */
+struct pru_rproc {
+ int id;
+ struct device *dev;
+ struct pruss *pruss;
+ struct rproc *rproc;
+ struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
+ const char *fw_name;
+};
+
+static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
+{
+ return readl_relaxed(pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
+}
+
+static inline
+void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
+{
+ writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
+}
+
+static int pru_rproc_start(struct rproc *rproc)
+{
+ struct device *dev = &rproc->dev;
+ struct pru_rproc *pru = rproc->priv;
+ u32 val;
+
+ dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
+ pru->id, (rproc->bootaddr >> 2));
+
+ val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
+ pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
+
+ return 0;
+}
+
+static int pru_rproc_stop(struct rproc *rproc)
+{
+ struct device *dev = &rproc->dev;
+ struct pru_rproc *pru = rproc->priv;
+ u32 val;
+
+ dev_dbg(dev, "stopping PRU%d\n", pru->id);
+
+ val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
+ val &= ~CTRL_CTRL_EN;
+ pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
+
+ return 0;
+}
+
+/*
+ * Convert PRU device address (data spaces only) to kernel virtual address.
+ *
+ * Each PRU has access to all data memories within the PRUSS, accessible at
+ * different ranges. So, look through both its primary and secondary Data
+ * RAMs as well as any shared Data RAM to convert a PRU device address to
+ * kernel virtual address. Data RAM0 is primary Data RAM for PRU0 and Data
+ * RAM1 is primary Data RAM for PRU1.
+ */
+static void *pru_d_da_to_va(struct pru_rproc *pru, u32 da, int len)
+{
+ struct pruss_mem_region dram0, dram1, shrd_ram;
+ struct pruss *pruss = pru->pruss;
+ u32 offset;
+ void *va = NULL;
+
+ if (len <= 0)
+ return NULL;
+
+ dram0 = pruss->mem_regions[PRUSS_MEM_DRAM0];
+ dram1 = pruss->mem_regions[PRUSS_MEM_DRAM1];
+ /* PRU1 has its local RAM addresses reversed */
+ if (pru->id == 1)
+ swap(dram0, dram1);
+ shrd_ram = pruss->mem_regions[PRUSS_MEM_SHRD_RAM2];
+
+ if (da >= PRU_PDRAM_DA && da + len <= PRU_PDRAM_DA + dram0.size) {
+ offset = da - PRU_PDRAM_DA;
+ va = (__force void *)(dram0.va + offset);
+ } else if (da >= PRU_SDRAM_DA &&
+ da + len <= PRU_SDRAM_DA + dram1.size) {
+ offset = da - PRU_SDRAM_DA;
+ va = (__force void *)(dram1.va + offset);
+ } else if (da >= PRU_SHRDRAM_DA &&
+ da + len <= PRU_SHRDRAM_DA + shrd_ram.size) {
+ offset = da - PRU_SHRDRAM_DA;
+ va = (__force void *)(shrd_ram.va + offset);
+ }
+
+ return va;
+}
+
+/*
+ * Convert PRU device address (instruction space) to kernel virtual address.
+ *
+ * A PRU does not have an unified address space. Each PRU has its very own
+ * private Instruction RAM, and its device address is identical to that of
+ * its primary Data RAM device address.
+ */
+static void *pru_i_da_to_va(struct pru_rproc *pru, u32 da, int len)
+{
+ u32 offset;
+ void *va = NULL;
+
+ if (len <= 0)
+ return NULL;
+
+ if (da >= PRU_IRAM_DA &&
+ da + len <= PRU_IRAM_DA + pru->mem_regions[PRU_IOMEM_IRAM].size) {
+ offset = da - PRU_IRAM_DA;
+ va = (__force void *)(pru->mem_regions[PRU_IOMEM_IRAM].va +
+ offset);
+ }
+
+ return va;
+}
+
+/*
+ * Provide address translations for only PRU Data RAMs through the remoteproc
+ * core for any PRU client drivers. The PRU Instruction RAM access is restricted
+ * only to the PRU loader code.
+ */
+static void *pru_rproc_da_to_va(struct rproc *rproc, u64 da, size_t len)
+{
+ struct pru_rproc *pru = rproc->priv;
+
+ return pru_d_da_to_va(pru, da, len);
+}
+
+/* PRU-specific address translator used by PRU loader. */
+static void *pru_da_to_va(struct rproc *rproc, u64 da, int len, bool is_iram)
+{
+ struct pru_rproc *pru = rproc->priv;
+ void *va;
+
+ if (is_iram)
+ va = pru_i_da_to_va(pru, da, len);
+ else
+ va = pru_d_da_to_va(pru, da, len);
+
+ return va;
+}
+
+static struct rproc_ops pru_rproc_ops = {
+ .start = pru_rproc_start,
+ .stop = pru_rproc_stop,
+ .da_to_va = pru_rproc_da_to_va,
+};
+
+static int
+pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
+{
+ struct device *dev = &rproc->dev;
+ struct elf32_hdr *ehdr;
+ struct elf32_phdr *phdr;
+ int i, ret = 0;
+ const u8 *elf_data = fw->data;
+
+ ehdr = (struct elf32_hdr *)elf_data;
+ phdr = (struct elf32_phdr *)(elf_data + ehdr->e_phoff);
+
+ /* go through the available ELF segments */
+ for (i = 0; i < ehdr->e_phnum; i++, phdr++) {
+ u32 da = phdr->p_paddr;
+ u32 memsz = phdr->p_memsz;
+ u32 filesz = phdr->p_filesz;
+ u32 offset = phdr->p_offset;
+ bool is_iram;
+ void *ptr;
+
+ if (phdr->p_type != PT_LOAD)
+ continue;
+
+ dev_dbg(dev, "phdr: type %d da 0x%x memsz 0x%x filesz 0x%x\n",
+ phdr->p_type, da, memsz, filesz);
+
+ if (filesz > memsz) {
+ dev_err(dev, "bad phdr filesz 0x%x memsz 0x%x\n",
+ filesz, memsz);
+ ret = -EINVAL;
+ break;
+ }
+
+ if (offset + filesz > fw->size) {
+ dev_err(dev, "truncated fw: need 0x%x avail 0x%zx\n",
+ offset + filesz, fw->size);
+ ret = -EINVAL;
+ break;
+ }
+
+ /* grab the kernel address for this device address */
+ is_iram = phdr->p_flags & PF_X;
+ ptr = pru_da_to_va(rproc, da, memsz, is_iram);
+ if (!ptr) {
+ dev_err(dev, "bad phdr da 0x%x mem 0x%x\n", da, memsz);
+ ret = -EINVAL;
+ break;
+ }
+
+ /* skip the memzero logic performed by remoteproc ELF loader */
+ if (!phdr->p_filesz)
+ continue;
+
+ memcpy(ptr, elf_data + phdr->p_offset, filesz);
+ }
+
+ return ret;
+}
+
+static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
+{
+ int ret;
+
+ /* load optional rsc table */
+ ret = rproc_elf_load_rsc_table(rproc, fw);
+ if (ret == -EINVAL)
+ dev_dbg(&rproc->dev, "no resource table found for this fw\n");
+
+ return ret;
+}
+
+/*
+ * Compute PRU id based on the IRAM addresses. The PRU IRAMs are
+ * always at a particular offset within the PRUSS address space.
+ */
+static int pru_rproc_set_id(struct pru_rproc *pru)
+{
+ int ret = 0;
+
+ switch (pru->mem_regions[PRU_IOMEM_IRAM].pa & 0x3ffff) {
+ case PRU0_IRAM_ADDR_MASK:
+ pru->id = 0;
+ break;
+ case PRU1_IRAM_ADDR_MASK:
+ pru->id = 1;
+ break;
+ default:
+ ret = -EINVAL;
+ }
+
+ return ret;
+}
+
+static int pru_rproc_probe(struct platform_device *pdev)
+{
+ struct device *dev = &pdev->dev;
+ struct device_node *np = dev->of_node;
+ struct platform_device *ppdev = to_platform_device(dev->parent);
+ struct pru_rproc *pru;
+ const char *fw_name;
+ struct rproc *rproc = NULL;
+ struct resource *res;
+ int i, ret;
+ const char *mem_names[PRU_IOMEM_MAX] = { "iram", "control", "debug" };
+
+ ret = of_property_read_string(np, "firmware-name", &fw_name);
+ if (ret) {
+ dev_err(dev, "unable to retrieve firmware-name %d\n", ret);
+ return ret;
+ }
+
+ rproc = devm_rproc_alloc(dev, pdev->name, &pru_rproc_ops, fw_name,
+ sizeof(*pru));
+ if (!rproc) {
+ dev_err(dev, "rproc_alloc failed\n");
+ return -ENOMEM;
+ }
+ /* use a custom load function to deal with PRU-specific quirks */
+ rproc->ops->load = pru_rproc_load_elf_segments;
+
+ /* use a custom parse function to deal with PRU-specific resources */
+ rproc->ops->parse_fw = pru_rproc_parse_fw;
+
+ /* error recovery is not supported for PRUs */
+ rproc->recovery_disabled = true;
+
+ /*
+ * rproc_add will auto-boot the processor normally, but this is not
+ * desired with PRU client driven boot-flow methodology. A PRU
+ * application/client driver will boot the corresponding PRU
+ * remote-processor as part of its state machine either through the
+ * remoteproc sysfs interface or through the equivalent kernel API.
+ */
+ rproc->auto_boot = false;
+
+ pru = rproc->priv;
+ pru->dev = dev;
+ pru->pruss = platform_get_drvdata(ppdev);
+ pru->rproc = rproc;
+ pru->fw_name = fw_name;
+
+ for (i = 0; i < ARRAY_SIZE(mem_names); i++) {
+ res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
+ mem_names[i]);
+ pru->mem_regions[i].va = devm_ioremap_resource(dev, res);
+ if (IS_ERR(pru->mem_regions[i].va)) {
+ dev_err(dev, "failed to parse and map memory resource %d %s\n",
+ i, mem_names[i]);
+ ret = PTR_ERR(pru->mem_regions[i].va);
+ return ret;
+ }
+ pru->mem_regions[i].pa = res->start;
+ pru->mem_regions[i].size = resource_size(res);
+
+ dev_dbg(dev, "memory %8s: pa %pa size 0x%zx va %pK\n",
+ mem_names[i], &pru->mem_regions[i].pa,
+ pru->mem_regions[i].size, pru->mem_regions[i].va);
+ }
+
+ ret = pru_rproc_set_id(pru);
+ if (ret < 0)
+ return ret;
+
+ platform_set_drvdata(pdev, rproc);
+
+ ret = devm_rproc_add(dev, pru->rproc);
+ if (ret) {
+ dev_err(dev, "rproc_add failed: %d\n", ret);
+ return ret;
+ }
+
+ dev_dbg(dev, "PRU rproc node %pOF probed successfully\n", np);
+
+ return 0;
+}
+
+static int pru_rproc_remove(struct platform_device *pdev)
+{
+ struct device *dev = &pdev->dev;
+ struct rproc *rproc = platform_get_drvdata(pdev);
+
+ dev_dbg(dev, "%s: removing rproc %s\n", __func__, rproc->name);
+
+ return 0;
+}
+
+static const struct of_device_id pru_rproc_match[] = {
+ { .compatible = "ti,am3356-pru", },
+ { .compatible = "ti,am4376-pru", },
+ { .compatible = "ti,am5728-pru", },
+ { .compatible = "ti,k2g-pru", },
+ {},
+};
+MODULE_DEVICE_TABLE(of, pru_rproc_match);
+
+static struct platform_driver pru_rproc_driver = {
+ .driver = {
+ .name = "pru-rproc",
+ .of_match_table = pru_rproc_match,
+ .suppress_bind_attrs = true,
+ },
+ .probe = pru_rproc_probe,
+ .remove = pru_rproc_remove,
+};
+module_platform_driver(pru_rproc_driver);
+
+MODULE_AUTHOR("Suman Anna <[email protected]>");
+MODULE_AUTHOR("Andrew F. Davis <[email protected]>");
+MODULE_AUTHOR("Grzegorz Jaszczyk <[email protected]>");
+MODULE_DESCRIPTION("PRU-ICSS Remote Processor Driver");
+MODULE_LICENSE("GPL v2");
--
2.29.0
From: Suman Anna <[email protected]>
The K3 J721E family of SoCs have a revised version of the AM65x ICSSG IP
and contains two instances of this newer ICSSG IP. Each ICSSG processor
subsystem contains 2 primary PRU cores, 2 auxiliary PRU cores called RTUs,
and 2 new auxiliary cores called Transmit PRUs (Tx_PRUs).
Enhance the existing PRU remoteproc driver to support these new PRU
and RTU cores by using specific compatibles. The cores have the same
memory copying limitations as on AM65x, so reuses the custom memcpy
function within the driver's ELF loader implementation. The initial
names for the firmware images for each PRU core are retrieved from
DT nodes, and can be adjusted through sysfs if required.
Signed-off-by: Suman Anna <[email protected]>
Co-developed-by: Grzegorz Jaszczyk <[email protected]>
Signed-off-by: Grzegorz Jaszczyk <[email protected]>
---
drivers/remoteproc/pru_rproc.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
index 04c9f07799e2..98f9c598993f 100644
--- a/drivers/remoteproc/pru_rproc.c
+++ b/drivers/remoteproc/pru_rproc.c
@@ -855,6 +855,9 @@ static const struct of_device_id pru_rproc_match[] = {
{ .compatible = "ti,am654-pru", .data = &k3_pru_data },
{ .compatible = "ti,am654-rtu", .data = &k3_rtu_data },
{ .compatible = "ti,am654-tx-pru", .data = &k3_tx_pru_data },
+ { .compatible = "ti,j721e-pru", .data = &k3_pru_data },
+ { .compatible = "ti,j721e-rtu", .data = &k3_rtu_data },
+ { .compatible = "ti,j721e-tx-pru", .data = &k3_tx_pru_data },
{},
};
MODULE_DEVICE_TABLE(of, pru_rproc_match);
--
2.29.0
From: Suman Anna <[email protected]>
The K3 AM65x family of SoCs have the next generation of the PRU-ICSS
processor subsystem, commonly referred to as ICSSG. Each ICSSG processor
subsystem on AM65x SR1.0 contains two primary PRU cores and two new
auxiliary PRU cores called RTUs. The AM65x SR2.0 SoCs have a revised
ICSSG IP that is based off the subsequent IP revision used on J721E
SoCs. This IP instance has two new custom auxiliary PRU cores called
Transmit PRUs (Tx_PRUs) in addition to the existing PRUs and RTUs.
Each RTU and Tx_PRU cores have their own dedicated IRAM (smaller than
a PRU), Control and debug feature sets, but is different in terms of
sub-modules integrated around it and does not have the full capabilities
associated with a PRU core. The RTU core is typically used to aid a
PRU core in accelerating data transfers, while the Tx_PRU cores is
normally used to control the TX L2 FIFO if enabled in Ethernet
applications. Both can also be used to run independent applications.
The RTU and Tx_PRU cores though share the same Data RAMs as the PRU
cores, so the memories have to be partitioned carefully between different
applications. The new cores also support a new sub-module called Task
Manager to support two different context thread executions.
Enhance the existing PRU remoteproc driver to support these new PRU, RTU
and Tx PRU cores by using specific compatibles. The initial names for the
firmware images for each PRU core are retrieved from DT nodes, and can
be adjusted through sysfs if required.
The PRU remoteproc driver has to be specifically modified to use a
custom memcpy function within its ELF loader implementation for these
new cores in order to overcome a limitation with copying data into each
of the core's IRAM memories. These memory ports support only 4-byte
writes, and any sub-word order byte writes clear out the remaining
bytes other than the bytes being written within the containing word.
The default ARM64 memcpy also cannot be used as it throws an exception
when the preferred 8-byte copy operation is attempted. This choice is
made by using a state flag that is set only on K3 SoCs.
Signed-off-by: Suman Anna <[email protected]>
Co-developed-by: Grzegorz Jaszczyk <[email protected]>
Signed-off-by: Grzegorz Jaszczyk <[email protected]>
---
drivers/remoteproc/pru_rproc.c | 141 ++++++++++++++++++++++++++++++---
1 file changed, 132 insertions(+), 9 deletions(-)
diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
index 33806ddcbd5d..04c9f07799e2 100644
--- a/drivers/remoteproc/pru_rproc.c
+++ b/drivers/remoteproc/pru_rproc.c
@@ -46,9 +46,13 @@
#define PRU_DEBUG_GPREG(x) (0x0000 + (x) * 4)
#define PRU_DEBUG_CT_REG(x) (0x0080 + (x) * 4)
-/* PRU Core IRAM address masks */
+/* PRU/RTU/Tx_PRU Core IRAM address masks */
#define PRU0_IRAM_ADDR_MASK 0x34000
#define PRU1_IRAM_ADDR_MASK 0x38000
+#define RTU0_IRAM_ADDR_MASK 0x4000
+#define RTU1_IRAM_ADDR_MASK 0x6000
+#define TX_PRU0_IRAM_ADDR_MASK 0xa000
+#define TX_PRU1_IRAM_ADDR_MASK 0xc000
/* PRU device addresses for various type of PRU RAMs */
#define PRU_IRAM_DA 0 /* Instruction RAM */
@@ -73,12 +77,38 @@ enum pru_iomem {
PRU_IOMEM_MAX,
};
+/**
+ * enum pru_type - PRU core type identifier
+ *
+ * @PRU_TYPE_PRU: Programmable Real-time Unit
+ * @PRU_TYPE_RTU: Auxiliary Programmable Real-Time Unit
+ * @PRU_TYPE_TX_PRU: Transmit Programmable Real-Time Unit
+ * @PRU_TYPE_MAX: just keep this one at the end
+ */
+enum pru_type {
+ PRU_TYPE_PRU = 0,
+ PRU_TYPE_RTU,
+ PRU_TYPE_TX_PRU,
+ PRU_TYPE_MAX,
+};
+
+/**
+ * struct pru_private_data - device data for a PRU core
+ * @type: type of the PRU core (PRU, RTU, Tx_PRU)
+ * @is_k3: flag used to identify the need for special load & event handling
+ */
+struct pru_private_data {
+ enum pru_type type;
+ unsigned int is_k3 : 1;
+};
+
/**
* struct pru_rproc - PRU remoteproc structure
* @id: id of the PRU core within the PRUSS
* @dev: PRU core device pointer
* @pruss: back-reference to parent PRUSS structure
* @rproc: remoteproc pointer for this PRU core
+ * @data: PRU core specific data
* @mem_regions: data for each of the PRU memory regions
* @fw_name: name of firmware image used during loading
* @mapped_irq: virtual interrupt numbers of created fw specific mapping
@@ -93,6 +123,7 @@ struct pru_rproc {
struct device *dev;
struct pruss *pruss;
struct rproc *rproc;
+ const struct pru_private_data *data;
struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
const char *fw_name;
int *mapped_irq;
@@ -318,11 +349,12 @@ static int pru_rproc_start(struct rproc *rproc)
{
struct device *dev = &rproc->dev;
struct pru_rproc *pru = rproc->priv;
+ const char *names[PRU_TYPE_MAX] = { "PRU", "RTU", "Tx_PRU" };
u32 val;
int ret;
- dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
- pru->id, (rproc->bootaddr >> 2));
+ dev_dbg(dev, "starting %s%d: entry-point = 0x%llx\n",
+ names[pru->data->type], pru->id, (rproc->bootaddr >> 2));
ret = pru_handle_intrmap(rproc);
/*
@@ -344,9 +376,10 @@ static int pru_rproc_stop(struct rproc *rproc)
{
struct device *dev = &rproc->dev;
struct pru_rproc *pru = rproc->priv;
+ const char *names[PRU_TYPE_MAX] = { "PRU", "RTU", "Tx_PRU" };
u32 val;
- dev_dbg(dev, "stopping PRU%d\n", pru->id);
+ dev_dbg(dev, "stopping %s%d\n", names[pru->data->type], pru->id);
val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
val &= ~CTRL_CTRL_EN;
@@ -458,9 +491,53 @@ static struct rproc_ops pru_rproc_ops = {
.da_to_va = pru_rproc_da_to_va,
};
+/*
+ * Custom memory copy implementation for ICSSG PRU/RTU Cores
+ *
+ * The ICSSG PRU/RTU cores have a memory copying issue with IRAM memories, that
+ * is not seen on previous generation SoCs. The data is reflected properly in
+ * the IRAM memories only for integer (4-byte) copies. Any unaligned copies
+ * result in all the other pre-existing bytes zeroed out within that 4-byte
+ * boundary, thereby resulting in wrong text/code in the IRAMs. Also, the
+ * IRAM memory port interface does not allow any 8-byte copies (as commonly
+ * used by ARM64 memcpy implementation) and throws an exception. The DRAM
+ * memory ports do not show this behavior. Use this custom copying function
+ * to properly load the PRU/RTU firmware images on all memories for simplicity.
+ */
+static int pru_rproc_memcpy(void *dest, const void *src, size_t count)
+{
+ const int *s = src;
+ int *d = dest;
+ int size = count / 4;
+ int *tmp_src = NULL;
+
+ /*
+ * TODO: relax limitation of 4-byte aligned dest addresses and copy
+ * sizes
+ */
+ if ((long)dest % 4 || count % 4)
+ return -EINVAL;
+
+ /* src offsets in ELF firmware image can be non-aligned */
+ if ((long)src % 4) {
+ tmp_src = kmemdup(src, count, GFP_KERNEL);
+ if (!tmp_src)
+ return -ENOMEM;
+ s = tmp_src;
+ }
+
+ while (size--)
+ *d++ = *s++;
+
+ kfree(tmp_src);
+
+ return 0;
+}
+
static int
pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
{
+ struct pru_rproc *pru = rproc->priv;
struct device *dev = &rproc->dev;
struct elf32_hdr *ehdr;
struct elf32_phdr *phdr;
@@ -512,7 +589,17 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
if (!phdr->p_filesz)
continue;
- memcpy(ptr, elf_data + phdr->p_offset, filesz);
+ if (pru->data->is_k3 && is_iram) {
+ ret = pru_rproc_memcpy(ptr, elf_data + phdr->p_offset,
+ filesz);
+ if (ret) {
+ dev_err(dev, "PRU memory copy failed for da 0x%x memsz 0x%x\n",
+ da, memsz);
+ break;
+ }
+ } else {
+ memcpy(ptr, elf_data + phdr->p_offset, filesz);
+ }
}
return ret;
@@ -619,9 +706,17 @@ static int pru_rproc_set_id(struct pru_rproc *pru)
int ret = 0;
switch (pru->mem_regions[PRU_IOMEM_IRAM].pa & 0x3ffff) {
+ case TX_PRU0_IRAM_ADDR_MASK:
+ fallthrough;
+ case RTU0_IRAM_ADDR_MASK:
+ fallthrough;
case PRU0_IRAM_ADDR_MASK:
pru->id = 0;
break;
+ case TX_PRU1_IRAM_ADDR_MASK:
+ fallthrough;
+ case RTU1_IRAM_ADDR_MASK:
+ fallthrough;
case PRU1_IRAM_ADDR_MASK:
pru->id = 1;
break;
@@ -642,8 +737,13 @@ static int pru_rproc_probe(struct platform_device *pdev)
struct rproc *rproc = NULL;
struct resource *res;
int i, ret;
+ const struct pru_private_data *data;
const char *mem_names[PRU_IOMEM_MAX] = { "iram", "control", "debug" };
+ data = of_device_get_match_data(&pdev->dev);
+ if (!data)
+ return -ENODEV;
+
ret = of_property_read_string(np, "firmware-name", &fw_name);
if (ret) {
dev_err(dev, "unable to retrieve firmware-name %d\n", ret);
@@ -676,6 +776,7 @@ static int pru_rproc_probe(struct platform_device *pdev)
pru = rproc->priv;
pru->dev = dev;
+ pru->data = data;
pru->pruss = platform_get_drvdata(ppdev);
pru->rproc = rproc;
pru->fw_name = fw_name;
@@ -727,11 +828,33 @@ static int pru_rproc_remove(struct platform_device *pdev)
return 0;
}
+static const struct pru_private_data pru_data = {
+ .type = PRU_TYPE_PRU,
+};
+
+static const struct pru_private_data k3_pru_data = {
+ .type = PRU_TYPE_PRU,
+ .is_k3 = 1,
+};
+
+static const struct pru_private_data k3_rtu_data = {
+ .type = PRU_TYPE_RTU,
+ .is_k3 = 1,
+};
+
+static const struct pru_private_data k3_tx_pru_data = {
+ .type = PRU_TYPE_TX_PRU,
+ .is_k3 = 1,
+};
+
static const struct of_device_id pru_rproc_match[] = {
- { .compatible = "ti,am3356-pru", },
- { .compatible = "ti,am4376-pru", },
- { .compatible = "ti,am5728-pru", },
- { .compatible = "ti,k2g-pru", },
+ { .compatible = "ti,am3356-pru", .data = &pru_data },
+ { .compatible = "ti,am4376-pru", .data = &pru_data },
+ { .compatible = "ti,am5728-pru", .data = &pru_data },
+ { .compatible = "ti,k2g-pru", .data = &pru_data },
+ { .compatible = "ti,am654-pru", .data = &k3_pru_data },
+ { .compatible = "ti,am654-rtu", .data = &k3_rtu_data },
+ { .compatible = "ti,am654-tx-pru", .data = &k3_tx_pru_data },
{},
};
MODULE_DEVICE_TABLE(of, pru_rproc_match);
--
2.29.0
From: Suman Anna <[email protected]>
The remoteproc core creates certain standard debugfs entries,
that does not give a whole lot of useful information for the
PRUs. The PRU remoteproc driver is enhanced to add additional
debugfs entries for PRU. These will be auto-cleaned up when
the parent rproc debug directory is removed.
The enhanced debugfs support adds two new entries: 'regs' and
'single_step'. The 'regs' dumps out the useful CTRL sub-module
registers as well as each of the 32 GPREGs and CT_REGs registers.
The GPREGs and CT_REGs though are printed only when the PRU is
halted and accessible as per the IP design.
The 'single_step' utilizes the single-step execution of the PRU
cores. Writing a non-zero value performs a single step, and a
zero value restores the PRU to execute in the same mode as the
mode before the first single step. (note: if the PRU is halted
because of a halt instruction, then no change occurs).
Logic for setting the PC and jumping over a halt instruction shall
be added in the future.
Signed-off-by: Suman Anna <[email protected]>
Signed-off-by: Grzegorz Jaszczyk <[email protected]>
---
drivers/remoteproc/pru_rproc.c | 137 +++++++++++++++++++++++++++++++++
1 file changed, 137 insertions(+)
diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
index 825e9c7e081b..33806ddcbd5d 100644
--- a/drivers/remoteproc/pru_rproc.c
+++ b/drivers/remoteproc/pru_rproc.c
@@ -11,6 +11,7 @@
*/
#include <linux/bitops.h>
+#include <linux/debugfs.h>
#include <linux/irqdomain.h>
#include <linux/module.h>
#include <linux/of_device.h>
@@ -25,6 +26,13 @@
/* PRU_ICSS_PRU_CTRL registers */
#define PRU_CTRL_CTRL 0x0000
#define PRU_CTRL_STS 0x0004
+#define PRU_CTRL_WAKEUP_EN 0x0008
+#define PRU_CTRL_CYCLE 0x000C
+#define PRU_CTRL_STALL 0x0010
+#define PRU_CTRL_CTBIR0 0x0020
+#define PRU_CTRL_CTBIR1 0x0024
+#define PRU_CTRL_CTPPR0 0x0028
+#define PRU_CTRL_CTPPR1 0x002C
/* CTRL register bit-fields */
#define CTRL_CTRL_SOFT_RST_N BIT(0)
@@ -34,6 +42,10 @@
#define CTRL_CTRL_SINGLE_STEP BIT(8)
#define CTRL_CTRL_RUNSTATE BIT(15)
+/* PRU_ICSS_PRU_DEBUG registers */
+#define PRU_DEBUG_GPREG(x) (0x0000 + (x) * 4)
+#define PRU_DEBUG_CT_REG(x) (0x0080 + (x) * 4)
+
/* PRU Core IRAM address masks */
#define PRU0_IRAM_ADDR_MASK 0x34000
#define PRU1_IRAM_ADDR_MASK 0x38000
@@ -73,6 +85,8 @@ enum pru_iomem {
* @pru_interrupt_map: pointer to interrupt mapping description (firmware)
* @pru_interrupt_map_sz: pru_interrupt_map size
* @evt_count: number of mapped events
+ * @dbg_single_step: debug state variable to set PRU into single step mode
+ * @dbg_continuous: debug state variable to restore PRU execution mode
*/
struct pru_rproc {
int id;
@@ -85,6 +99,8 @@ struct pru_rproc {
struct pru_irq_rsc *pru_interrupt_map;
size_t pru_interrupt_map_sz;
ssize_t evt_count;
+ u32 dbg_single_step;
+ u32 dbg_continuous;
};
static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
@@ -98,6 +114,125 @@ void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
}
+static inline u32 pru_debug_read_reg(struct pru_rproc *pru, unsigned int reg)
+{
+ return readl_relaxed(pru->mem_regions[PRU_IOMEM_DEBUG].va + reg);
+}
+
+static int regs_show(struct seq_file *s, void *data)
+{
+ struct rproc *rproc = s->private;
+ struct pru_rproc *pru = rproc->priv;
+ int i, nregs = 32;
+ u32 pru_sts;
+ int pru_is_running;
+
+ seq_puts(s, "============== Control Registers ==============\n");
+ seq_printf(s, "CTRL := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_CTRL));
+ pru_sts = pru_control_read_reg(pru, PRU_CTRL_STS);
+ seq_printf(s, "STS (PC) := 0x%08x (0x%08x)\n", pru_sts, pru_sts << 2);
+ seq_printf(s, "WAKEUP_EN := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_WAKEUP_EN));
+ seq_printf(s, "CYCLE := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_CYCLE));
+ seq_printf(s, "STALL := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_STALL));
+ seq_printf(s, "CTBIR0 := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_CTBIR0));
+ seq_printf(s, "CTBIR1 := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_CTBIR1));
+ seq_printf(s, "CTPPR0 := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_CTPPR0));
+ seq_printf(s, "CTPPR1 := 0x%08x\n",
+ pru_control_read_reg(pru, PRU_CTRL_CTPPR1));
+
+ seq_puts(s, "=============== Debug Registers ===============\n");
+ pru_is_running = pru_control_read_reg(pru, PRU_CTRL_CTRL) &
+ CTRL_CTRL_RUNSTATE;
+ if (pru_is_running) {
+ seq_puts(s, "PRU is executing, cannot print/access debug registers.\n");
+ return 0;
+ }
+
+ for (i = 0; i < nregs; i++) {
+ seq_printf(s, "GPREG%-2d := 0x%08x\tCT_REG%-2d := 0x%08x\n",
+ i, pru_debug_read_reg(pru, PRU_DEBUG_GPREG(i)),
+ i, pru_debug_read_reg(pru, PRU_DEBUG_CT_REG(i)));
+ }
+
+ return 0;
+}
+
+DEFINE_SHOW_ATTRIBUTE(regs);
+
+/*
+ * Control PRU single-step mode
+ *
+ * This is a debug helper function used for controlling the single-step
+ * mode of the PRU. The PRU Debug registers are not accessible when the
+ * PRU is in RUNNING state.
+ *
+ * Writing a non-zero value sets the PRU into single-step mode irrespective
+ * of its previous state. The PRU mode is saved only on the first set into
+ * a single-step mode. Writing a zero value will restore the PRU into its
+ * original mode.
+ */
+static int pru_rproc_debug_ss_set(void *data, u64 val)
+{
+ struct rproc *rproc = data;
+ struct pru_rproc *pru = rproc->priv;
+ u32 reg_val;
+
+ val = val ? 1 : 0;
+ if (!val && !pru->dbg_single_step)
+ return 0;
+
+ reg_val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
+
+ if (val && !pru->dbg_single_step)
+ pru->dbg_continuous = reg_val;
+
+ if (val)
+ reg_val |= CTRL_CTRL_SINGLE_STEP | CTRL_CTRL_EN;
+ else
+ reg_val = pru->dbg_continuous;
+
+ pru->dbg_single_step = val;
+ pru_control_write_reg(pru, PRU_CTRL_CTRL, reg_val);
+
+ return 0;
+}
+
+static int pru_rproc_debug_ss_get(void *data, u64 *val)
+{
+ struct rproc *rproc = data;
+ struct pru_rproc *pru = rproc->priv;
+
+ *val = pru->dbg_single_step;
+
+ return 0;
+}
+DEFINE_SIMPLE_ATTRIBUTE(pru_rproc_debug_ss_fops, pru_rproc_debug_ss_get,
+ pru_rproc_debug_ss_set, "%llu\n");
+
+/*
+ * Create PRU-specific debugfs entries
+ *
+ * The entries are created only if the parent remoteproc debugfs directory
+ * exists, and will be cleaned up by the remoteproc core.
+ */
+static void pru_rproc_create_debug_entries(struct rproc *rproc)
+{
+ if (!rproc->dbg_dir)
+ return;
+
+ debugfs_create_file("regs", 0400, rproc->dbg_dir,
+ rproc, ®s_fops);
+ debugfs_create_file("single_step", 0600, rproc->dbg_dir,
+ rproc, &pru_rproc_debug_ss_fops);
+}
+
static void pru_dispose_irq_mapping(struct pru_rproc *pru)
{
while (--pru->evt_count >= 0) {
@@ -575,6 +710,8 @@ static int pru_rproc_probe(struct platform_device *pdev)
return ret;
}
+ pru_rproc_create_debug_entries(rproc);
+
dev_dbg(dev, "PRU rproc node %pOF probed successfully\n", np);
return 0;
--
2.29.0
The firmware blob can contain optional ELF sections: .resource_table
section and .pru_irq_map one. The second one contains the PRUSS
interrupt mapping description, which needs to be setup before powering
on the PRU core. To avoid RAM wastage this ELF section is not mapped to
any ELF segment (by the firmware linker) and therefore is not loaded to
PRU memory.
The PRU interrupt configuration is handled within the PRUSS INTC irqchip
driver and leverages the system events to interrupt channels and host
interrupts mapping configuration. Relevant irq routing information is
passed through a special .pru_irq_map ELF section (for interrupts routed
to and used by PRU cores) or via the PRU application's device tree node
(for interrupts routed to and used by the main CPU). The mappings are
currently programmed during the booting/shutdown of the PRU.
The interrupt configuration passed through .pru_irq_map ELF section is
optional. It varies on specific firmware functionality and therefore
have to be unwinded during PRU stop and performed again during
PRU start.
Co-developed-by: Suman Anna <[email protected]>
Signed-off-by: Suman Anna <[email protected]>
Signed-off-by: Grzegorz Jaszczyk <[email protected]>
---
drivers/remoteproc/pru_rproc.c | 191 ++++++++++++++++++++++++++++++++-
drivers/remoteproc/pru_rproc.h | 46 ++++++++
2 files changed, 236 insertions(+), 1 deletion(-)
create mode 100644 drivers/remoteproc/pru_rproc.h
diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
index c94c8e965c21..825e9c7e081b 100644
--- a/drivers/remoteproc/pru_rproc.c
+++ b/drivers/remoteproc/pru_rproc.c
@@ -11,13 +11,16 @@
*/
#include <linux/bitops.h>
+#include <linux/irqdomain.h>
#include <linux/module.h>
#include <linux/of_device.h>
+#include <linux/of_irq.h>
#include <linux/pruss_driver.h>
#include <linux/remoteproc.h>
#include "remoteproc_internal.h"
#include "remoteproc_elf_helpers.h"
+#include "pru_rproc.h"
/* PRU_ICSS_PRU_CTRL registers */
#define PRU_CTRL_CTRL 0x0000
@@ -41,6 +44,8 @@
#define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
#define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
+#define MAX_PRU_SYS_EVENTS 160
+
/**
* enum pru_iomem - PRU core memory/register range identifiers
*
@@ -64,6 +69,10 @@ enum pru_iomem {
* @rproc: remoteproc pointer for this PRU core
* @mem_regions: data for each of the PRU memory regions
* @fw_name: name of firmware image used during loading
+ * @mapped_irq: virtual interrupt numbers of created fw specific mapping
+ * @pru_interrupt_map: pointer to interrupt mapping description (firmware)
+ * @pru_interrupt_map_sz: pru_interrupt_map size
+ * @evt_count: number of mapped events
*/
struct pru_rproc {
int id;
@@ -72,6 +81,10 @@ struct pru_rproc {
struct rproc *rproc;
struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
const char *fw_name;
+ int *mapped_irq;
+ struct pru_irq_rsc *pru_interrupt_map;
+ size_t pru_interrupt_map_sz;
+ ssize_t evt_count;
};
static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
@@ -85,15 +98,107 @@ void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
}
+static void pru_dispose_irq_mapping(struct pru_rproc *pru)
+{
+ while (--pru->evt_count >= 0) {
+ if (pru->mapped_irq[pru->evt_count] > 0)
+ irq_dispose_mapping(pru->mapped_irq[pru->evt_count]);
+ }
+
+ kfree(pru->mapped_irq);
+}
+
+/*
+ * Parse the custom PRU interrupt map resource and configure the INTC
+ * appropriately.
+ */
+static int pru_handle_intrmap(struct rproc *rproc)
+{
+ struct device *dev = rproc->dev.parent;
+ struct pru_rproc *pru = rproc->priv;
+ struct pru_irq_rsc *rsc = pru->pru_interrupt_map;
+ struct irq_fwspec fwspec;
+ struct device_node *irq_parent;
+ int i, ret = 0;
+
+ /* not having pru_interrupt_map is not an error */
+ if (!rsc)
+ return 0;
+
+ /* currently supporting only type 0 */
+ if (rsc->type != 0) {
+ dev_err(dev, "unsupported rsc type: %d\n", rsc->type);
+ return -EINVAL;
+ }
+
+ if (rsc->num_evts < 0 || rsc->num_evts > MAX_PRU_SYS_EVENTS)
+ return -EINVAL;
+
+ if (sizeof(*rsc) + rsc->num_evts * sizeof(struct pruss_int_map) !=
+ pru->pru_interrupt_map_sz)
+ return -EINVAL;
+
+ pru->evt_count = rsc->num_evts;
+ pru->mapped_irq = kcalloc(pru->evt_count, sizeof(int), GFP_KERNEL);
+ if (!pru->mapped_irq)
+ return -ENOMEM;
+
+ /*
+ * parse and fill in system event to interrupt channel and
+ * channel-to-host mapping
+ */
+ irq_parent = of_irq_find_parent(pru->dev->of_node);
+ if (!irq_parent) {
+ kfree(pru->mapped_irq);
+ return -ENODEV;
+ }
+
+ fwspec.fwnode = of_node_to_fwnode(irq_parent);
+ fwspec.param_count = 3;
+ for (i = 0; i < pru->evt_count; i++) {
+ fwspec.param[0] = rsc->pru_intc_map[i].event;
+ fwspec.param[1] = rsc->pru_intc_map[i].chnl;
+ fwspec.param[2] = rsc->pru_intc_map[i].host;
+
+ dev_dbg(dev, "mapping%d: event %d, chnl %d, host %d\n",
+ i, fwspec.param[0], fwspec.param[1], fwspec.param[2]);
+
+ pru->mapped_irq[i] = irq_create_fwspec_mapping(&fwspec);
+ if (pru->mapped_irq[i] < 0) {
+ dev_err(dev, "failed to get virq\n");
+ ret = pru->mapped_irq[i];
+ goto map_fail;
+ }
+ }
+
+ return ret;
+
+map_fail:
+ pru_dispose_irq_mapping(pru);
+
+ return ret;
+}
+
static int pru_rproc_start(struct rproc *rproc)
{
struct device *dev = &rproc->dev;
struct pru_rproc *pru = rproc->priv;
u32 val;
+ int ret;
dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
pru->id, (rproc->bootaddr >> 2));
+ ret = pru_handle_intrmap(rproc);
+ /*
+ * reset references to pru interrupt map - they will stop being valid
+ * after rproc_start returns
+ */
+ pru->pru_interrupt_map = NULL;
+ pru->pru_interrupt_map_sz = 0;
+ if (ret)
+ return ret;
+
val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
@@ -112,6 +217,10 @@ static int pru_rproc_stop(struct rproc *rproc)
val &= ~CTRL_CTRL_EN;
pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
+ /* dispose irq mapping - new firmware can provide new mapping */
+ if (pru->mapped_irq)
+ pru_dispose_irq_mapping(pru);
+
return 0;
}
@@ -274,16 +383,96 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
return ret;
}
+static const void *
+pru_rproc_find_interrupt_map(struct device *dev, const struct firmware *fw)
+{
+ const void *shdr, *name_table_shdr;
+ const char *name_table;
+ const u8 *elf_data = (void *)fw->data;
+ u8 class = fw_elf_get_class(fw);
+ size_t fw_size = fw->size;
+ const void *ehdr = elf_data;
+ u16 shnum = elf_hdr_get_e_shnum(class, ehdr);
+ u32 elf_shdr_get_size = elf_size_of_shdr(class);
+ u16 shstrndx = elf_hdr_get_e_shstrndx(class, ehdr);
+ int i;
+
+ /* first, get the section header according to the elf class */
+ shdr = elf_data + elf_hdr_get_e_shoff(class, ehdr);
+ /* compute name table section header entry in shdr array */
+ name_table_shdr = shdr + (shstrndx * elf_shdr_get_size);
+ /* finally, compute the name table section address in elf */
+ name_table = elf_data + elf_shdr_get_sh_offset(class, name_table_shdr);
+
+ for (i = 0; i < shnum; i++, shdr += elf_shdr_get_size) {
+ u64 size = elf_shdr_get_sh_size(class, shdr);
+ u64 offset = elf_shdr_get_sh_offset(class, shdr);
+ u32 name = elf_shdr_get_sh_name(class, shdr);
+
+ if (strcmp(name_table + name, ".pru_irq_map"))
+ continue;
+
+ /* make sure we have the entire table */
+ if (offset + size > fw_size || offset + size < size) {
+ dev_err(dev, "interrupt map sec truncated\n");
+ return ERR_PTR(-EINVAL);
+ }
+
+ /* make sure table has at least the header */
+ if (sizeof(struct pru_irq_rsc) > size) {
+ dev_err(dev, "header-less interrupt map sec\n");
+ return ERR_PTR(-EINVAL);
+ }
+
+ return shdr;
+ }
+
+ dev_dbg(dev, "no .pru_irq_map section found for this fw\n");
+
+ return NULL;
+}
+
+/*
+ * Usa a custom parse_fw callback function for dealing with standard
+ * resource table and a PRU-specific custom ELF section.
+ *
+ * The firmware blob can contain optional ELF sections: .resource_table section
+ * and .pru_irq_map one. The second one contains the PRUSS interrupt mapping
+ * description, which needs to be setup before powering on the PRU core. To
+ * avoid RAM wastage this ELF section is not mapped to any ELF segment (by the
+ * firmware linker) and therefore is not loaded to PRU memory.
+ */
static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
{
+ struct device *dev = &rproc->dev;
+ struct pru_rproc *pru = rproc->priv;
+ const u8 *elf_data = fw->data;
+ const void *shdr;
+ u8 class = fw_elf_get_class(fw);
+ u64 sh_offset;
int ret;
/* load optional rsc table */
ret = rproc_elf_load_rsc_table(rproc, fw);
if (ret == -EINVAL)
dev_dbg(&rproc->dev, "no resource table found for this fw\n");
+ else if (ret)
+ return ret;
- return ret;
+ /* find .pru_interrupt_map section, not having it is not an error */
+ shdr = pru_rproc_find_interrupt_map(dev, fw);
+ if (IS_ERR(shdr))
+ return PTR_ERR(shdr);
+
+ if (!shdr)
+ return 0;
+
+ /* preserve pointer to PRU interrupt map together with it size */
+ sh_offset = elf_shdr_get_sh_offset(class, shdr);
+ pru->pru_interrupt_map = (struct pru_irq_rsc *)(elf_data + sh_offset);
+ pru->pru_interrupt_map_sz = elf_shdr_get_sh_size(class, shdr);
+
+ return 0;
}
/*
diff --git a/drivers/remoteproc/pru_rproc.h b/drivers/remoteproc/pru_rproc.h
new file mode 100644
index 000000000000..8ee9c3171610
--- /dev/null
+++ b/drivers/remoteproc/pru_rproc.h
@@ -0,0 +1,46 @@
+/* SPDX-License-Identifier: (GPL-2.0-only OR BSD-3-Clause) */
+/*
+ * PRUSS Remote Processor specific types
+ *
+ * Copyright (C) 2014-2020 Texas Instruments Incorporated - https://www.ti.com/
+ * Suman Anna <[email protected]>
+ */
+
+#ifndef _PRU_RPROC_H_
+#define _PRU_RPROC_H_
+
+/**
+ * struct pruss_int_map - PRU system events _to_ channel and host mapping
+ * @event: number of the system event
+ * @chnl: channel number assigned to a given @event
+ * @host: host number assigned to a given @chnl
+ *
+ * PRU system events are mapped to channels, and these channels are mapped
+ * to host interrupts. Events can be mapped to channels in a one-to-one or
+ * many-to-one ratio (multiple events per channel), and channels can be
+ * mapped to host interrupts in a one-to-one or many-to-one ratio (multiple
+ * channels per interrupt).
+ */
+struct pruss_int_map {
+ u8 event;
+ u8 chnl;
+ u8 host;
+};
+
+/**
+ * struct pru_irq_rsc - PRU firmware section header for IRQ data
+ * @type: resource type
+ * @num_evts: number of described events
+ * @pru_intc_map: PRU interrupt routing description
+ *
+ * The PRU firmware blob can contain optional .pru_irq_map ELF section, which
+ * provides the PRUSS interrupt mapping description. The pru_irq_rsc struct
+ * describes resource entry format.
+ */
+struct pru_irq_rsc {
+ u8 type;
+ u8 num_evts;
+ struct pruss_int_map pru_intc_map[];
+} __packed;
+
+#endif /* _PRU_RPROC_H_ */
--
2.29.0
Hi Greg,
I have a few minor comments below..
On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> From: Suman Anna <[email protected]>
>
> The Programmable Real-Time Unit Subsystem (PRUSS) consists of
> dual 32-bit RISC cores (Programmable Real-Time Units, or PRUs)
> for program execution. This patch adds a remoteproc platform
> driver for managing the individual PRU RISC cores life cycle.
>
> The PRUs do not have a unified address space (have an Instruction
> RAM and a primary Data RAM at both 0x0). The PRU remoteproc driver
> therefore uses a custom remoteproc core ELF loader ops. The added
> .da_to_va ops is only used to provide translations for the PRU
> Data RAMs. This remoteproc driver does not have support for error
> recovery and system suspend/resume features. Different compatibles
> are used to allow providing scalability for instance-specific device
> data if needed. The driver uses a default firmware-name retrieved
> from device-tree for each PRU core, and the firmwares are expected
> to be present in the standard Linux firmware search paths. They can
> also be adjusted by userspace if required through the sysfs interface
> provided by the remoteproc core.
>
> The PRU remoteproc driver uses a client-driven boot methodology: it
> does _not_ support auto-boot so that the PRU load and boot is dictated
> by the corresponding client drivers for achieving various usecases.
> This allows flexibility for the client drivers or applications to set
> a firmware name (if needed) based on their desired functionality and
> boot the PRU. The sysfs bind and unbind attributes have also been
> suppressed so that the PRU devices cannot be unbound and thereby
> shutdown a PRU from underneath a PRU client driver.
>
> The driver currently supports the AM335x, AM437x, AM57xx and 66AK2G
> SoCs, and support for other TI SoCs will be added in subsequent
> patches.
>
> Co-developed-by: Andrew F. Davis <[email protected]>
> Signed-off-by: Andrew F. Davis <[email protected]>
> Signed-off-by: Suman Anna <[email protected]>
> Co-developed-by: Grzegorz Jaszczyk <[email protected]>
> Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> ---
> drivers/remoteproc/Kconfig | 12 +
> drivers/remoteproc/Makefile | 1 +
> drivers/remoteproc/pru_rproc.c | 428 +++++++++++++++++++++++++++++++++
> 3 files changed, 441 insertions(+)
> create mode 100644 drivers/remoteproc/pru_rproc.c
>
> diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig
> index d99548fb5dde..3e3865a7cd78 100644
> --- a/drivers/remoteproc/Kconfig
> +++ b/drivers/remoteproc/Kconfig
> @@ -125,6 +125,18 @@ config KEYSTONE_REMOTEPROC
> It's safe to say N here if you're not interested in the Keystone
> DSPs or just want to use a bare minimum kernel.
>
> +config PRU_REMOTEPROC
> + tristate "TI PRU remoteproc support"
> + depends on TI_PRUSS
> + default TI_PRUSS
> + help
> + Support for TI PRU remote processors present within a PRU-ICSS
> + subsystem via the remote processor framework.
> +
> + Say Y or M here to support the Programmable Realtime Unit (PRU)
> + processors on various TI SoCs. It's safe to say N here if you're
> + not interested in the PRU or if you are unsure.
> +
> config QCOM_PIL_INFO
> tristate
>
> diff --git a/drivers/remoteproc/Makefile b/drivers/remoteproc/Makefile
> index da2ace4ec86c..bb26c9e4ef9c 100644
> --- a/drivers/remoteproc/Makefile
> +++ b/drivers/remoteproc/Makefile
> @@ -18,6 +18,7 @@ obj-$(CONFIG_OMAP_REMOTEPROC) += omap_remoteproc.o
> obj-$(CONFIG_WKUP_M3_RPROC) += wkup_m3_rproc.o
> obj-$(CONFIG_DA8XX_REMOTEPROC) += da8xx_remoteproc.o
> obj-$(CONFIG_KEYSTONE_REMOTEPROC) += keystone_remoteproc.o
> +obj-$(CONFIG_PRU_REMOTEPROC) += pru_rproc.o
> obj-$(CONFIG_QCOM_PIL_INFO) += qcom_pil_info.o
> obj-$(CONFIG_QCOM_RPROC_COMMON) += qcom_common.o
> obj-$(CONFIG_QCOM_Q6V5_COMMON) += qcom_q6v5.o
> diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> new file mode 100644
> index 000000000000..c94c8e965c21
> --- /dev/null
> +++ b/drivers/remoteproc/pru_rproc.c
> @@ -0,0 +1,428 @@
> +// SPDX-License-Identifier: GPL-2.0-only
> +/*
> + * PRU-ICSS remoteproc driver for various TI SoCs
> + *
> + * Copyright (C) 2014-2020 Texas Instruments Incorporated - https://www.ti.com/
> + *
> + * Author(s):
> + * Suman Anna <[email protected]>
> + * Andrew F. Davis <[email protected]>
> + * Grzegorz Jaszczyk <[email protected]> for Texas Instruments
> + */
> +
> +#include <linux/bitops.h>
> +#include <linux/module.h>
> +#include <linux/of_device.h>
> +#include <linux/pruss_driver.h>
> +#include <linux/remoteproc.h>
> +
> +#include "remoteproc_internal.h"
> +#include "remoteproc_elf_helpers.h"
> +
> +/* PRU_ICSS_PRU_CTRL registers */
> +#define PRU_CTRL_CTRL 0x0000
> +#define PRU_CTRL_STS 0x0004
> +
> +/* CTRL register bit-fields */
> +#define CTRL_CTRL_SOFT_RST_N BIT(0)
> +#define CTRL_CTRL_EN BIT(1)
> +#define CTRL_CTRL_SLEEPING BIT(2)
> +#define CTRL_CTRL_CTR_EN BIT(3)
> +#define CTRL_CTRL_SINGLE_STEP BIT(8)
> +#define CTRL_CTRL_RUNSTATE BIT(15)
> +
> +/* PRU Core IRAM address masks */
> +#define PRU0_IRAM_ADDR_MASK 0x34000
> +#define PRU1_IRAM_ADDR_MASK 0x38000
> +
> +/* PRU device addresses for various type of PRU RAMs */
> +#define PRU_IRAM_DA 0 /* Instruction RAM */
> +#define PRU_PDRAM_DA 0 /* Primary Data RAM */
> +#define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
> +#define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
> +
> +/**
> + * enum pru_iomem - PRU core memory/register range identifiers
> + *
> + * @PRU_IOMEM_IRAM: PRU Instruction RAM range
> + * @PRU_IOMEM_CTRL: PRU Control register range
> + * @PRU_IOMEM_DEBUG: PRU Debug register range
> + * @PRU_IOMEM_MAX: just keep this one at the end
> + */
> +enum pru_iomem {
> + PRU_IOMEM_IRAM = 0,
> + PRU_IOMEM_CTRL,
> + PRU_IOMEM_DEBUG,
> + PRU_IOMEM_MAX,
> +};
> +
> +/**
> + * struct pru_rproc - PRU remoteproc structure
> + * @id: id of the PRU core within the PRUSS
> + * @dev: PRU core device pointer
> + * @pruss: back-reference to parent PRUSS structure
> + * @rproc: remoteproc pointer for this PRU core
> + * @mem_regions: data for each of the PRU memory regions
> + * @fw_name: name of firmware image used during loading
> + */
> +struct pru_rproc {
> + int id;
> + struct device *dev;
> + struct pruss *pruss;
> + struct rproc *rproc;
> + struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> + const char *fw_name;
> +};
> +
> +static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
> +{
> + return readl_relaxed(pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> +}
> +
> +static inline
> +void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
> +{
> + writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> +}
> +
> +static int pru_rproc_start(struct rproc *rproc)
> +{
> + struct device *dev = &rproc->dev;
> + struct pru_rproc *pru = rproc->priv;
> + u32 val;
> +
> + dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> + pru->id, (rproc->bootaddr >> 2));
> +
> + val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
> + pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> +
> + return 0;
> +}
> +
> +static int pru_rproc_stop(struct rproc *rproc)
> +{
> + struct device *dev = &rproc->dev;
> + struct pru_rproc *pru = rproc->priv;
> + u32 val;
> +
> + dev_dbg(dev, "stopping PRU%d\n", pru->id);
> +
> + val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
> + val &= ~CTRL_CTRL_EN;
> + pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> +
> + return 0;
> +}
> +
> +/*
> + * Convert PRU device address (data spaces only) to kernel virtual address.
> + *
> + * Each PRU has access to all data memories within the PRUSS, accessible at
> + * different ranges. So, look through both its primary and secondary Data
> + * RAMs as well as any shared Data RAM to convert a PRU device address to
> + * kernel virtual address. Data RAM0 is primary Data RAM for PRU0 and Data
> + * RAM1 is primary Data RAM for PRU1.
> + */
> +static void *pru_d_da_to_va(struct pru_rproc *pru, u32 da, int len)
The .da_to_va() ops argument has changed to using size_t in recent kernels, so
let's convert the type from int to size_t for the len argument.
> +{
> + struct pruss_mem_region dram0, dram1, shrd_ram;
> + struct pruss *pruss = pru->pruss;
> + u32 offset;
> + void *va = NULL;
> +
> + if (len <= 0)
> + return NULL;
And we adjust this accordingly.
> +
> + dram0 = pruss->mem_regions[PRUSS_MEM_DRAM0];
> + dram1 = pruss->mem_regions[PRUSS_MEM_DRAM1];
> + /* PRU1 has its local RAM addresses reversed */
> + if (pru->id == 1)
> + swap(dram0, dram1);
> + shrd_ram = pruss->mem_regions[PRUSS_MEM_SHRD_RAM2];
> +
> + if (da >= PRU_PDRAM_DA && da + len <= PRU_PDRAM_DA + dram0.size) {
> + offset = da - PRU_PDRAM_DA;
> + va = (__force void *)(dram0.va + offset);
> + } else if (da >= PRU_SDRAM_DA &&
> + da + len <= PRU_SDRAM_DA + dram1.size) {
> + offset = da - PRU_SDRAM_DA;
> + va = (__force void *)(dram1.va + offset);
> + } else if (da >= PRU_SHRDRAM_DA &&
> + da + len <= PRU_SHRDRAM_DA + shrd_ram.size) {
> + offset = da - PRU_SHRDRAM_DA;
> + va = (__force void *)(shrd_ram.va + offset);
> + }
> +
> + return va;
> +}
> +
> +/*
> + * Convert PRU device address (instruction space) to kernel virtual address.
> + *
> + * A PRU does not have an unified address space. Each PRU has its very own
> + * private Instruction RAM, and its device address is identical to that of
> + * its primary Data RAM device address.
> + */
> +static void *pru_i_da_to_va(struct pru_rproc *pru, u32 da, int len)
> +{
> + u32 offset;
> + void *va = NULL;
> +
> + if (len <= 0)
> + return NULL;
Same comments as above.
> +
> + if (da >= PRU_IRAM_DA &&
> + da + len <= PRU_IRAM_DA + pru->mem_regions[PRU_IOMEM_IRAM].size) {
> + offset = da - PRU_IRAM_DA;
> + va = (__force void *)(pru->mem_regions[PRU_IOMEM_IRAM].va +
> + offset);
> + }
> +
> + return va;
> +}
> +
> +/*
> + * Provide address translations for only PRU Data RAMs through the remoteproc
> + * core for any PRU client drivers. The PRU Instruction RAM access is restricted
> + * only to the PRU loader code.
> + */
> +static void *pru_rproc_da_to_va(struct rproc *rproc, u64 da, size_t len)
> +{
> + struct pru_rproc *pru = rproc->priv;
> +
> + return pru_d_da_to_va(pru, da, len);
> +}
> +
> +/* PRU-specific address translator used by PRU loader. */
> +static void *pru_da_to_va(struct rproc *rproc, u64 da, int len, bool is_iram)
> +{
> + struct pru_rproc *pru = rproc->priv;
> + void *va;
> +
> + if (is_iram)
> + va = pru_i_da_to_va(pru, da, len);
> + else
> + va = pru_d_da_to_va(pru, da, len);
> +
> + return va;
> +}
> +
> +static struct rproc_ops pru_rproc_ops = {
> + .start = pru_rproc_start,
> + .stop = pru_rproc_stop,
> + .da_to_va = pru_rproc_da_to_va,
> +};
> +
> +static int
> +pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> +{
> + struct device *dev = &rproc->dev;
> + struct elf32_hdr *ehdr;
> + struct elf32_phdr *phdr;
> + int i, ret = 0;
> + const u8 *elf_data = fw->data;
> +
> + ehdr = (struct elf32_hdr *)elf_data;
> + phdr = (struct elf32_phdr *)(elf_data + ehdr->e_phoff);
> +
> + /* go through the available ELF segments */
> + for (i = 0; i < ehdr->e_phnum; i++, phdr++) {
> + u32 da = phdr->p_paddr;
> + u32 memsz = phdr->p_memsz;
> + u32 filesz = phdr->p_filesz;
> + u32 offset = phdr->p_offset;
> + bool is_iram;
> + void *ptr;
> +
> + if (phdr->p_type != PT_LOAD)
> + continue;
> +
> + dev_dbg(dev, "phdr: type %d da 0x%x memsz 0x%x filesz 0x%x\n",
> + phdr->p_type, da, memsz, filesz);
> +
> + if (filesz > memsz) {
> + dev_err(dev, "bad phdr filesz 0x%x memsz 0x%x\n",
> + filesz, memsz);
> + ret = -EINVAL;
> + break;
> + }
> +
> + if (offset + filesz > fw->size) {
> + dev_err(dev, "truncated fw: need 0x%x avail 0x%zx\n",
> + offset + filesz, fw->size);
> + ret = -EINVAL;
> + break;
> + }
> +
> + /* grab the kernel address for this device address */
> + is_iram = phdr->p_flags & PF_X;
> + ptr = pru_da_to_va(rproc, da, memsz, is_iram);
> + if (!ptr) {
> + dev_err(dev, "bad phdr da 0x%x mem 0x%x\n", da, memsz);
> + ret = -EINVAL;
> + break;
> + }
> +
> + /* skip the memzero logic performed by remoteproc ELF loader */
> + if (!phdr->p_filesz)
> + continue;
> +
> + memcpy(ptr, elf_data + phdr->p_offset, filesz);
> + }
> +
> + return ret;
> +}
> +
> +static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
> +{
> + int ret;
> +
> + /* load optional rsc table */
> + ret = rproc_elf_load_rsc_table(rproc, fw);
> + if (ret == -EINVAL)
> + dev_dbg(&rproc->dev, "no resource table found for this fw\n");
> +
> + return ret;
This should return 0 if there is no resource table, otherwise this fails. You
have this corrected in Patch 3, but please move that logic here.
> +}
> +
> +/*
> + * Compute PRU id based on the IRAM addresses. The PRU IRAMs are
> + * always at a particular offset within the PRUSS address space.
> + */
> +static int pru_rproc_set_id(struct pru_rproc *pru)
> +{
> + int ret = 0;
> +
> + switch (pru->mem_regions[PRU_IOMEM_IRAM].pa & 0x3ffff) {
We can probably update this to use a macro for 0x3fffff.
regards
Suman
> + case PRU0_IRAM_ADDR_MASK:
> + pru->id = 0;
> + break;
> + case PRU1_IRAM_ADDR_MASK:
> + pru->id = 1;
> + break;
> + default:
> + ret = -EINVAL;
> + }
> +
> + return ret;
> +}
> +
> +static int pru_rproc_probe(struct platform_device *pdev)
> +{
> + struct device *dev = &pdev->dev;
> + struct device_node *np = dev->of_node;
> + struct platform_device *ppdev = to_platform_device(dev->parent);
> + struct pru_rproc *pru;
> + const char *fw_name;
> + struct rproc *rproc = NULL;
> + struct resource *res;
> + int i, ret;
> + const char *mem_names[PRU_IOMEM_MAX] = { "iram", "control", "debug" };
> +
> + ret = of_property_read_string(np, "firmware-name", &fw_name);
> + if (ret) {
> + dev_err(dev, "unable to retrieve firmware-name %d\n", ret);
> + return ret;
> + }
> +
> + rproc = devm_rproc_alloc(dev, pdev->name, &pru_rproc_ops, fw_name,
> + sizeof(*pru));
> + if (!rproc) {
> + dev_err(dev, "rproc_alloc failed\n");
> + return -ENOMEM;
> + }
> + /* use a custom load function to deal with PRU-specific quirks */
> + rproc->ops->load = pru_rproc_load_elf_segments;
> +
> + /* use a custom parse function to deal with PRU-specific resources */
> + rproc->ops->parse_fw = pru_rproc_parse_fw;
> +
> + /* error recovery is not supported for PRUs */
> + rproc->recovery_disabled = true;
> +
> + /*
> + * rproc_add will auto-boot the processor normally, but this is not
> + * desired with PRU client driven boot-flow methodology. A PRU
> + * application/client driver will boot the corresponding PRU
> + * remote-processor as part of its state machine either through the
> + * remoteproc sysfs interface or through the equivalent kernel API.
> + */
> + rproc->auto_boot = false;
> +
> + pru = rproc->priv;
> + pru->dev = dev;
> + pru->pruss = platform_get_drvdata(ppdev);
> + pru->rproc = rproc;
> + pru->fw_name = fw_name;
> +
> + for (i = 0; i < ARRAY_SIZE(mem_names); i++) {
> + res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
> + mem_names[i]);
> + pru->mem_regions[i].va = devm_ioremap_resource(dev, res);
> + if (IS_ERR(pru->mem_regions[i].va)) {
> + dev_err(dev, "failed to parse and map memory resource %d %s\n",
> + i, mem_names[i]);
> + ret = PTR_ERR(pru->mem_regions[i].va);
> + return ret;
> + }
> + pru->mem_regions[i].pa = res->start;
> + pru->mem_regions[i].size = resource_size(res);
> +
> + dev_dbg(dev, "memory %8s: pa %pa size 0x%zx va %pK\n",
> + mem_names[i], &pru->mem_regions[i].pa,
> + pru->mem_regions[i].size, pru->mem_regions[i].va);
> + }
> +
> + ret = pru_rproc_set_id(pru);
> + if (ret < 0)
> + return ret;
> +
> + platform_set_drvdata(pdev, rproc);
> +
> + ret = devm_rproc_add(dev, pru->rproc);
> + if (ret) {
> + dev_err(dev, "rproc_add failed: %d\n", ret);
> + return ret;
> + }
> +
> + dev_dbg(dev, "PRU rproc node %pOF probed successfully\n", np);
> +
> + return 0;
> +}
> +
> +static int pru_rproc_remove(struct platform_device *pdev)
> +{
> + struct device *dev = &pdev->dev;
> + struct rproc *rproc = platform_get_drvdata(pdev);
> +
> + dev_dbg(dev, "%s: removing rproc %s\n", __func__, rproc->name);
> +
> + return 0;
> +}
> +
> +static const struct of_device_id pru_rproc_match[] = {
> + { .compatible = "ti,am3356-pru", },
> + { .compatible = "ti,am4376-pru", },
> + { .compatible = "ti,am5728-pru", },
> + { .compatible = "ti,k2g-pru", },
> + {},
> +};
> +MODULE_DEVICE_TABLE(of, pru_rproc_match);
> +
> +static struct platform_driver pru_rproc_driver = {
> + .driver = {
> + .name = "pru-rproc",
> + .of_match_table = pru_rproc_match,
> + .suppress_bind_attrs = true,
> + },
> + .probe = pru_rproc_probe,
> + .remove = pru_rproc_remove,
> +};
> +module_platform_driver(pru_rproc_driver);
> +
> +MODULE_AUTHOR("Suman Anna <[email protected]>");
> +MODULE_AUTHOR("Andrew F. Davis <[email protected]>");
> +MODULE_AUTHOR("Grzegorz Jaszczyk <[email protected]>");
> +MODULE_DESCRIPTION("PRU-ICSS Remote Processor Driver");
> +MODULE_LICENSE("GPL v2");
>
Hi Greg,
On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> From: Suman Anna <[email protected]>
>
> The K3 AM65x family of SoCs have the next generation of the PRU-ICSS
> processor subsystem, commonly referred to as ICSSG. Each ICSSG processor
> subsystem on AM65x SR1.0 contains two primary PRU cores and two new
> auxiliary PRU cores called RTUs. The AM65x SR2.0 SoCs have a revised
> ICSSG IP that is based off the subsequent IP revision used on J721E
> SoCs. This IP instance has two new custom auxiliary PRU cores called
> Transmit PRUs (Tx_PRUs) in addition to the existing PRUs and RTUs.
>
> Each RTU and Tx_PRU cores have their own dedicated IRAM (smaller than
> a PRU), Control and debug feature sets, but is different in terms of
> sub-modules integrated around it and does not have the full capabilities
> associated with a PRU core. The RTU core is typically used to aid a
> PRU core in accelerating data transfers, while the Tx_PRU cores is
> normally used to control the TX L2 FIFO if enabled in Ethernet
> applications. Both can also be used to run independent applications.
> The RTU and Tx_PRU cores though share the same Data RAMs as the PRU
> cores, so the memories have to be partitioned carefully between different
> applications. The new cores also support a new sub-module called Task
> Manager to support two different context thread executions.
>
> Enhance the existing PRU remoteproc driver to support these new PRU, RTU
> and Tx PRU cores by using specific compatibles. The initial names for the
> firmware images for each PRU core are retrieved from DT nodes, and can
> be adjusted through sysfs if required.
>
> The PRU remoteproc driver has to be specifically modified to use a
> custom memcpy function within its ELF loader implementation for these
> new cores in order to overcome a limitation with copying data into each
> of the core's IRAM memories. These memory ports support only 4-byte
> writes, and any sub-word order byte writes clear out the remaining
> bytes other than the bytes being written within the containing word.
> The default ARM64 memcpy also cannot be used as it throws an exception
> when the preferred 8-byte copy operation is attempted. This choice is
> made by using a state flag that is set only on K3 SoCs.
>
> Signed-off-by: Suman Anna <[email protected]>
> Co-developed-by: Grzegorz Jaszczyk <[email protected]>
> Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> ---
> drivers/remoteproc/pru_rproc.c | 141 ++++++++++++++++++++++++++++++---
> 1 file changed, 132 insertions(+), 9 deletions(-)
>
> diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> index 33806ddcbd5d..04c9f07799e2 100644
> --- a/drivers/remoteproc/pru_rproc.c
> +++ b/drivers/remoteproc/pru_rproc.c
> @@ -46,9 +46,13 @@
> #define PRU_DEBUG_GPREG(x) (0x0000 + (x) * 4)
> #define PRU_DEBUG_CT_REG(x) (0x0080 + (x) * 4)
>
> -/* PRU Core IRAM address masks */
> +/* PRU/RTU/Tx_PRU Core IRAM address masks */
> #define PRU0_IRAM_ADDR_MASK 0x34000
> #define PRU1_IRAM_ADDR_MASK 0x38000
> +#define RTU0_IRAM_ADDR_MASK 0x4000
> +#define RTU1_IRAM_ADDR_MASK 0x6000
> +#define TX_PRU0_IRAM_ADDR_MASK 0xa000
> +#define TX_PRU1_IRAM_ADDR_MASK 0xc000
>
> /* PRU device addresses for various type of PRU RAMs */
> #define PRU_IRAM_DA 0 /* Instruction RAM */
> @@ -73,12 +77,38 @@ enum pru_iomem {
> PRU_IOMEM_MAX,
> };
>
> +/**
> + * enum pru_type - PRU core type identifier
> + *
> + * @PRU_TYPE_PRU: Programmable Real-time Unit
> + * @PRU_TYPE_RTU: Auxiliary Programmable Real-Time Unit
> + * @PRU_TYPE_TX_PRU: Transmit Programmable Real-Time Unit
> + * @PRU_TYPE_MAX: just keep this one at the end
> + */
> +enum pru_type {
> + PRU_TYPE_PRU = 0,
> + PRU_TYPE_RTU,
> + PRU_TYPE_TX_PRU,
> + PRU_TYPE_MAX,
> +};
> +
> +/**
> + * struct pru_private_data - device data for a PRU core
> + * @type: type of the PRU core (PRU, RTU, Tx_PRU)
> + * @is_k3: flag used to identify the need for special load & event handling
> + */
> +struct pru_private_data {
> + enum pru_type type;
> + unsigned int is_k3 : 1;
> +};
> +
> /**
> * struct pru_rproc - PRU remoteproc structure
> * @id: id of the PRU core within the PRUSS
> * @dev: PRU core device pointer
> * @pruss: back-reference to parent PRUSS structure
> * @rproc: remoteproc pointer for this PRU core
> + * @data: PRU core specific data
> * @mem_regions: data for each of the PRU memory regions
> * @fw_name: name of firmware image used during loading
> * @mapped_irq: virtual interrupt numbers of created fw specific mapping
> @@ -93,6 +123,7 @@ struct pru_rproc {
> struct device *dev;
> struct pruss *pruss;
> struct rproc *rproc;
> + const struct pru_private_data *data;
> struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> const char *fw_name;
> int *mapped_irq;
> @@ -318,11 +349,12 @@ static int pru_rproc_start(struct rproc *rproc)
> {
> struct device *dev = &rproc->dev;
> struct pru_rproc *pru = rproc->priv;
> + const char *names[PRU_TYPE_MAX] = { "PRU", "RTU", "Tx_PRU" };
> u32 val;
> int ret;
>
> - dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> - pru->id, (rproc->bootaddr >> 2));
> + dev_dbg(dev, "starting %s%d: entry-point = 0x%llx\n",
> + names[pru->data->type], pru->id, (rproc->bootaddr >> 2));
>
> ret = pru_handle_intrmap(rproc);
> /*
> @@ -344,9 +376,10 @@ static int pru_rproc_stop(struct rproc *rproc)
> {
> struct device *dev = &rproc->dev;
> struct pru_rproc *pru = rproc->priv;
> + const char *names[PRU_TYPE_MAX] = { "PRU", "RTU", "Tx_PRU" };
> u32 val;
>
> - dev_dbg(dev, "stopping PRU%d\n", pru->id);
> + dev_dbg(dev, "stopping %s%d\n", names[pru->data->type], pru->id);
>
> val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
> val &= ~CTRL_CTRL_EN;
> @@ -458,9 +491,53 @@ static struct rproc_ops pru_rproc_ops = {
> .da_to_va = pru_rproc_da_to_va,
> };
>
> +/*
> + * Custom memory copy implementation for ICSSG PRU/RTU Cores
Please update this to add Tx_PRU as well to the list here and in the below
description.
> + *
> + * The ICSSG PRU/RTU cores have a memory copying issue with IRAM memories, that
> + * is not seen on previous generation SoCs. The data is reflected properly in
> + * the IRAM memories only for integer (4-byte) copies. Any unaligned copies
> + * result in all the other pre-existing bytes zeroed out within that 4-byte
> + * boundary, thereby resulting in wrong text/code in the IRAMs. Also, the
> + * IRAM memory port interface does not allow any 8-byte copies (as commonly
> + * used by ARM64 memcpy implementation) and throws an exception. The DRAM
> + * memory ports do not show this behavior. Use this custom copying function
> + * to properly load the PRU/RTU firmware images on all memories for simplicity.
This last line is obsolete now that we use regular memcpy for Data RAM copies.
regards
Suman
> + */
> +static int pru_rproc_memcpy(void *dest, const void *src, size_t count)
> +{
> + const int *s = src;
> + int *d = dest;
> + int size = count / 4;
> + int *tmp_src = NULL;
> +
> + /*
> + * TODO: relax limitation of 4-byte aligned dest addresses and copy
> + * sizes
> + */
> + if ((long)dest % 4 || count % 4)
> + return -EINVAL;
> +
> + /* src offsets in ELF firmware image can be non-aligned */
> + if ((long)src % 4) {
> + tmp_src = kmemdup(src, count, GFP_KERNEL);
> + if (!tmp_src)
> + return -ENOMEM;
> + s = tmp_src;
> + }
> +
> + while (size--)
> + *d++ = *s++;
> +
> + kfree(tmp_src);
> +
> + return 0;
> +}
> +
> static int
> pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> {
> + struct pru_rproc *pru = rproc->priv;
> struct device *dev = &rproc->dev;
> struct elf32_hdr *ehdr;
> struct elf32_phdr *phdr;
> @@ -512,7 +589,17 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> if (!phdr->p_filesz)
> continue;
>
> - memcpy(ptr, elf_data + phdr->p_offset, filesz);
> + if (pru->data->is_k3 && is_iram) {
> + ret = pru_rproc_memcpy(ptr, elf_data + phdr->p_offset,
> + filesz);
> + if (ret) {
> + dev_err(dev, "PRU memory copy failed for da 0x%x memsz 0x%x\n",
> + da, memsz);
> + break;
> + }
> + } else {
> + memcpy(ptr, elf_data + phdr->p_offset, filesz);
> + }
> }
>
> return ret;
> @@ -619,9 +706,17 @@ static int pru_rproc_set_id(struct pru_rproc *pru)
> int ret = 0;
>
> switch (pru->mem_regions[PRU_IOMEM_IRAM].pa & 0x3ffff) {
> + case TX_PRU0_IRAM_ADDR_MASK:
> + fallthrough;
> + case RTU0_IRAM_ADDR_MASK:
> + fallthrough;
> case PRU0_IRAM_ADDR_MASK:
> pru->id = 0;
> break;
> + case TX_PRU1_IRAM_ADDR_MASK:
> + fallthrough;
> + case RTU1_IRAM_ADDR_MASK:
> + fallthrough;
> case PRU1_IRAM_ADDR_MASK:
> pru->id = 1;
> break;
> @@ -642,8 +737,13 @@ static int pru_rproc_probe(struct platform_device *pdev)
> struct rproc *rproc = NULL;
> struct resource *res;
> int i, ret;
> + const struct pru_private_data *data;
> const char *mem_names[PRU_IOMEM_MAX] = { "iram", "control", "debug" };
>
> + data = of_device_get_match_data(&pdev->dev);
> + if (!data)
> + return -ENODEV;
> +
> ret = of_property_read_string(np, "firmware-name", &fw_name);
> if (ret) {
> dev_err(dev, "unable to retrieve firmware-name %d\n", ret);
> @@ -676,6 +776,7 @@ static int pru_rproc_probe(struct platform_device *pdev)
>
> pru = rproc->priv;
> pru->dev = dev;
> + pru->data = data;
> pru->pruss = platform_get_drvdata(ppdev);
> pru->rproc = rproc;
> pru->fw_name = fw_name;
> @@ -727,11 +828,33 @@ static int pru_rproc_remove(struct platform_device *pdev)
> return 0;
> }
>
> +static const struct pru_private_data pru_data = {
> + .type = PRU_TYPE_PRU,
> +};
> +
> +static const struct pru_private_data k3_pru_data = {
> + .type = PRU_TYPE_PRU,
> + .is_k3 = 1,
> +};
> +
> +static const struct pru_private_data k3_rtu_data = {
> + .type = PRU_TYPE_RTU,
> + .is_k3 = 1,
> +};
> +
> +static const struct pru_private_data k3_tx_pru_data = {
> + .type = PRU_TYPE_TX_PRU,
> + .is_k3 = 1,
> +};
> +
> static const struct of_device_id pru_rproc_match[] = {
> - { .compatible = "ti,am3356-pru", },
> - { .compatible = "ti,am4376-pru", },
> - { .compatible = "ti,am5728-pru", },
> - { .compatible = "ti,k2g-pru", },
> + { .compatible = "ti,am3356-pru", .data = &pru_data },
> + { .compatible = "ti,am4376-pru", .data = &pru_data },
> + { .compatible = "ti,am5728-pru", .data = &pru_data },
> + { .compatible = "ti,k2g-pru", .data = &pru_data },
> + { .compatible = "ti,am654-pru", .data = &k3_pru_data },
> + { .compatible = "ti,am654-rtu", .data = &k3_rtu_data },
> + { .compatible = "ti,am654-tx-pru", .data = &k3_tx_pru_data },
> {},
> };
> MODULE_DEVICE_TABLE(of, pru_rproc_match);
>
Hi Greg,
On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> The firmware blob can contain optional ELF sections: .resource_table
> section and .pru_irq_map one. The second one contains the PRUSS
> interrupt mapping description, which needs to be setup before powering
> on the PRU core. To avoid RAM wastage this ELF section is not mapped to
> any ELF segment (by the firmware linker) and therefore is not loaded to
> PRU memory.
>
> The PRU interrupt configuration is handled within the PRUSS INTC irqchip
> driver and leverages the system events to interrupt channels and host
> interrupts mapping configuration. Relevant irq routing information is
> passed through a special .pru_irq_map ELF section (for interrupts routed
> to and used by PRU cores) or via the PRU application's device tree node
> (for interrupts routed to and used by the main CPU). The mappings are
> currently programmed during the booting/shutdown of the PRU.
>
> The interrupt configuration passed through .pru_irq_map ELF section is
> optional. It varies on specific firmware functionality and therefore
> have to be unwinded during PRU stop and performed again during
> PRU start.
>
> Co-developed-by: Suman Anna <[email protected]>
> Signed-off-by: Suman Anna <[email protected]>
> Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> ---
> drivers/remoteproc/pru_rproc.c | 191 ++++++++++++++++++++++++++++++++-
> drivers/remoteproc/pru_rproc.h | 46 ++++++++
> 2 files changed, 236 insertions(+), 1 deletion(-)
> create mode 100644 drivers/remoteproc/pru_rproc.h
>
> diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> index c94c8e965c21..825e9c7e081b 100644
> --- a/drivers/remoteproc/pru_rproc.c
> +++ b/drivers/remoteproc/pru_rproc.c
> @@ -11,13 +11,16 @@
> */
>
> #include <linux/bitops.h>
> +#include <linux/irqdomain.h>
> #include <linux/module.h>
> #include <linux/of_device.h>
> +#include <linux/of_irq.h>
> #include <linux/pruss_driver.h>
> #include <linux/remoteproc.h>
>
> #include "remoteproc_internal.h"
> #include "remoteproc_elf_helpers.h"
> +#include "pru_rproc.h"
>
> /* PRU_ICSS_PRU_CTRL registers */
> #define PRU_CTRL_CTRL 0x0000
> @@ -41,6 +44,8 @@
> #define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
> #define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
>
> +#define MAX_PRU_SYS_EVENTS 160
> +
> /**
> * enum pru_iomem - PRU core memory/register range identifiers
> *
> @@ -64,6 +69,10 @@ enum pru_iomem {
> * @rproc: remoteproc pointer for this PRU core
> * @mem_regions: data for each of the PRU memory regions
> * @fw_name: name of firmware image used during loading
> + * @mapped_irq: virtual interrupt numbers of created fw specific mapping
> + * @pru_interrupt_map: pointer to interrupt mapping description (firmware)
> + * @pru_interrupt_map_sz: pru_interrupt_map size
> + * @evt_count: number of mapped events
> */
> struct pru_rproc {
> int id;
> @@ -72,6 +81,10 @@ struct pru_rproc {
> struct rproc *rproc;
> struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> const char *fw_name;
> + int *mapped_irq;
> + struct pru_irq_rsc *pru_interrupt_map;
> + size_t pru_interrupt_map_sz;
> + ssize_t evt_count;
Do you really need this to be ssize_t type?
> };
>
> static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
> @@ -85,15 +98,107 @@ void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
> writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> }
>
> +static void pru_dispose_irq_mapping(struct pru_rproc *pru)
> +{
> + while (--pru->evt_count >= 0) {
> + if (pru->mapped_irq[pru->evt_count] > 0)
> + irq_dispose_mapping(pru->mapped_irq[pru->evt_count]);
> + }
> +
> + kfree(pru->mapped_irq);
> +}
> +
> +/*
> + * Parse the custom PRU interrupt map resource and configure the INTC
> + * appropriately.
> + */
> +static int pru_handle_intrmap(struct rproc *rproc)
> +{
> + struct device *dev = rproc->dev.parent;
> + struct pru_rproc *pru = rproc->priv;
> + struct pru_irq_rsc *rsc = pru->pru_interrupt_map;
> + struct irq_fwspec fwspec;
> + struct device_node *irq_parent;
> + int i, ret = 0;
> +
> + /* not having pru_interrupt_map is not an error */
> + if (!rsc)
> + return 0;
> +
> + /* currently supporting only type 0 */
> + if (rsc->type != 0) {
> + dev_err(dev, "unsupported rsc type: %d\n", rsc->type);
> + return -EINVAL;
> + }
> +
> + if (rsc->num_evts < 0 || rsc->num_evts > MAX_PRU_SYS_EVENTS)
> + return -EINVAL;
> +
> + if (sizeof(*rsc) + rsc->num_evts * sizeof(struct pruss_int_map) !=
> + pru->pru_interrupt_map_sz)
> + return -EINVAL;
> +
> + pru->evt_count = rsc->num_evts;
> + pru->mapped_irq = kcalloc(pru->evt_count, sizeof(int), GFP_KERNEL);
> + if (!pru->mapped_irq)
> + return -ENOMEM;
> +
> + /*
> + * parse and fill in system event to interrupt channel and
> + * channel-to-host mapping
> + */
> + irq_parent = of_irq_find_parent(pru->dev->of_node);
> + if (!irq_parent) {
> + kfree(pru->mapped_irq);
> + return -ENODEV;
> + }
> +
> + fwspec.fwnode = of_node_to_fwnode(irq_parent);
> + fwspec.param_count = 3;
> + for (i = 0; i < pru->evt_count; i++) {
> + fwspec.param[0] = rsc->pru_intc_map[i].event;
> + fwspec.param[1] = rsc->pru_intc_map[i].chnl;
> + fwspec.param[2] = rsc->pru_intc_map[i].host;
> +
> + dev_dbg(dev, "mapping%d: event %d, chnl %d, host %d\n",
> + i, fwspec.param[0], fwspec.param[1], fwspec.param[2]);
> +
> + pru->mapped_irq[i] = irq_create_fwspec_mapping(&fwspec);
> + if (pru->mapped_irq[i] < 0) {
> + dev_err(dev, "failed to get virq\n");
> + ret = pru->mapped_irq[i];
> + goto map_fail;
> + }
> + }
> +
> + return ret;
> +
> +map_fail:
> + pru_dispose_irq_mapping(pru);
> +
> + return ret;
> +}
> +
> static int pru_rproc_start(struct rproc *rproc)
> {
> struct device *dev = &rproc->dev;
> struct pru_rproc *pru = rproc->priv;
> u32 val;
> + int ret;
>
> dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> pru->id, (rproc->bootaddr >> 2));
>
> + ret = pru_handle_intrmap(rproc);
> + /*
> + * reset references to pru interrupt map - they will stop being valid
> + * after rproc_start returns
> + */
> + pru->pru_interrupt_map = NULL;
> + pru->pru_interrupt_map_sz = 0;
> + if (ret)
> + return ret;
> +
> val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
> pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
>
> @@ -112,6 +217,10 @@ static int pru_rproc_stop(struct rproc *rproc)
> val &= ~CTRL_CTRL_EN;
> pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
>
> + /* dispose irq mapping - new firmware can provide new mapping */
> + if (pru->mapped_irq)
> + pru_dispose_irq_mapping(pru);
> +
> return 0;
> }
>
> @@ -274,16 +383,96 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> return ret;
> }
>
> +static const void *
> +pru_rproc_find_interrupt_map(struct device *dev, const struct firmware *fw)
> +{
> + const void *shdr, *name_table_shdr;
> + const char *name_table;
> + const u8 *elf_data = (void *)fw->data;
> + u8 class = fw_elf_get_class(fw);
> + size_t fw_size = fw->size;
> + const void *ehdr = elf_data;
> + u16 shnum = elf_hdr_get_e_shnum(class, ehdr);
> + u32 elf_shdr_get_size = elf_size_of_shdr(class);
> + u16 shstrndx = elf_hdr_get_e_shstrndx(class, ehdr);
> + int i;
> +
> + /* first, get the section header according to the elf class */
> + shdr = elf_data + elf_hdr_get_e_shoff(class, ehdr);
> + /* compute name table section header entry in shdr array */
> + name_table_shdr = shdr + (shstrndx * elf_shdr_get_size);
> + /* finally, compute the name table section address in elf */
> + name_table = elf_data + elf_shdr_get_sh_offset(class, name_table_shdr);
I see you used the style influenced by the remoteproc_elf_loader code. PRUs are
all 32-bit, so we need not use this strictly. I am ok with this style, but
prefer consistent usage style between this function and
pru_rproc_load_elf_segments().
> +
> + for (i = 0; i < shnum; i++, shdr += elf_shdr_get_size) {
> + u64 size = elf_shdr_get_sh_size(class, shdr);
> + u64 offset = elf_shdr_get_sh_offset(class, shdr);
> + u32 name = elf_shdr_get_sh_name(class, shdr);
> +
> + if (strcmp(name_table + name, ".pru_irq_map"))
> + continue;
> +
> + /* make sure we have the entire table */
> + if (offset + size > fw_size || offset + size < size) {
> + dev_err(dev, "interrupt map sec truncated\n");
sec can confuse developers, suggest rephrasing this trace, something like
".pru_irq_map section truncated"
> + return ERR_PTR(-EINVAL);
> + }
> +
> + /* make sure table has at least the header */
> + if (sizeof(struct pru_irq_rsc) > size) {
> + dev_err(dev, "header-less interrupt map sec\n");
same comment as above
> + return ERR_PTR(-EINVAL);
> + }
> +
> + return shdr;
> + }
> +
> + dev_dbg(dev, "no .pru_irq_map section found for this fw\n");
> +
> + return NULL;
> +}
> +
> +/*
> + * Usa a custom parse_fw callback function for dealing with standard
> + * resource table and a PRU-specific custom ELF section.
> + *
> + * The firmware blob can contain optional ELF sections: .resource_table section
> + * and .pru_irq_map one. The second one contains the PRUSS interrupt mapping
> + * description, which needs to be setup before powering on the PRU core. To
> + * avoid RAM wastage this ELF section is not mapped to any ELF segment (by the
> + * firmware linker) and therefore is not loaded to PRU memory.
Some of this description can move to Patch 2 as well.
> + */
> static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
> {
> + struct device *dev = &rproc->dev;
> + struct pru_rproc *pru = rproc->priv;
> + const u8 *elf_data = fw->data;
> + const void *shdr;
> + u8 class = fw_elf_get_class(fw);
> + u64 sh_offset;
> int ret;
>
> /* load optional rsc table */
> ret = rproc_elf_load_rsc_table(rproc, fw);
> if (ret == -EINVAL)
> dev_dbg(&rproc->dev, "no resource table found for this fw\n");
> + else if (ret)
> + return ret;
This hunk should be part of Patch 2.
>
> - return ret;
> + /* find .pru_interrupt_map section, not having it is not an error */
> + shdr = pru_rproc_find_interrupt_map(dev, fw);
> + if (IS_ERR(shdr))
> + return PTR_ERR(shdr);
> +
> + if (!shdr)
> + return 0;
> +
> + /* preserve pointer to PRU interrupt map together with it size */
> + sh_offset = elf_shdr_get_sh_offset(class, shdr);
> + pru->pru_interrupt_map = (struct pru_irq_rsc *)(elf_data + sh_offset);
> + pru->pru_interrupt_map_sz = elf_shdr_get_sh_size(class, shdr);
> +
> + return 0;
And this one as well.
regards
Suman
> }
>
> /*
> diff --git a/drivers/remoteproc/pru_rproc.h b/drivers/remoteproc/pru_rproc.h
> new file mode 100644
> index 000000000000..8ee9c3171610
> --- /dev/null
> +++ b/drivers/remoteproc/pru_rproc.h
> @@ -0,0 +1,46 @@
> +/* SPDX-License-Identifier: (GPL-2.0-only OR BSD-3-Clause) */
> +/*
> + * PRUSS Remote Processor specific types
> + *
> + * Copyright (C) 2014-2020 Texas Instruments Incorporated - https://www.ti.com/
> + * Suman Anna <[email protected]>
> + */
> +
> +#ifndef _PRU_RPROC_H_
> +#define _PRU_RPROC_H_
> +
> +/**
> + * struct pruss_int_map - PRU system events _to_ channel and host mapping
> + * @event: number of the system event
> + * @chnl: channel number assigned to a given @event
> + * @host: host number assigned to a given @chnl
> + *
> + * PRU system events are mapped to channels, and these channels are mapped
> + * to host interrupts. Events can be mapped to channels in a one-to-one or
> + * many-to-one ratio (multiple events per channel), and channels can be
> + * mapped to host interrupts in a one-to-one or many-to-one ratio (multiple
> + * channels per interrupt).
> + */
> +struct pruss_int_map {
> + u8 event;
> + u8 chnl;
> + u8 host;
> +};
> +
> +/**
> + * struct pru_irq_rsc - PRU firmware section header for IRQ data
> + * @type: resource type
> + * @num_evts: number of described events
> + * @pru_intc_map: PRU interrupt routing description
> + *
> + * The PRU firmware blob can contain optional .pru_irq_map ELF section, which
> + * provides the PRUSS interrupt mapping description. The pru_irq_rsc struct
> + * describes resource entry format.
> + */
> +struct pru_irq_rsc {
> + u8 type;
> + u8 num_evts;
> + struct pruss_int_map pru_intc_map[];
> +} __packed;
> +
> +#endif /* _PRU_RPROC_H_ */
>
Hi Suman,
On Tue, 17 Nov 2020 at 20:41, Suman Anna <[email protected]> wrote:
>
> Hi Greg,
>
> I have a few minor comments below..
>
> On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> > From: Suman Anna <[email protected]>
> >
> > The Programmable Real-Time Unit Subsystem (PRUSS) consists of
> > dual 32-bit RISC cores (Programmable Real-Time Units, or PRUs)
> > for program execution. This patch adds a remoteproc platform
> > driver for managing the individual PRU RISC cores life cycle.
> >
> > The PRUs do not have a unified address space (have an Instruction
> > RAM and a primary Data RAM at both 0x0). The PRU remoteproc driver
> > therefore uses a custom remoteproc core ELF loader ops. The added
> > .da_to_va ops is only used to provide translations for the PRU
> > Data RAMs. This remoteproc driver does not have support for error
> > recovery and system suspend/resume features. Different compatibles
> > are used to allow providing scalability for instance-specific device
> > data if needed. The driver uses a default firmware-name retrieved
> > from device-tree for each PRU core, and the firmwares are expected
> > to be present in the standard Linux firmware search paths. They can
> > also be adjusted by userspace if required through the sysfs interface
> > provided by the remoteproc core.
> >
> > The PRU remoteproc driver uses a client-driven boot methodology: it
> > does _not_ support auto-boot so that the PRU load and boot is dictated
> > by the corresponding client drivers for achieving various usecases.
> > This allows flexibility for the client drivers or applications to set
> > a firmware name (if needed) based on their desired functionality and
> > boot the PRU. The sysfs bind and unbind attributes have also been
> > suppressed so that the PRU devices cannot be unbound and thereby
> > shutdown a PRU from underneath a PRU client driver.
> >
> > The driver currently supports the AM335x, AM437x, AM57xx and 66AK2G
> > SoCs, and support for other TI SoCs will be added in subsequent
> > patches.
> >
> > Co-developed-by: Andrew F. Davis <[email protected]>
> > Signed-off-by: Andrew F. Davis <[email protected]>
> > Signed-off-by: Suman Anna <[email protected]>
> > Co-developed-by: Grzegorz Jaszczyk <[email protected]>
> > Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> > ---
> > drivers/remoteproc/Kconfig | 12 +
> > drivers/remoteproc/Makefile | 1 +
> > drivers/remoteproc/pru_rproc.c | 428 +++++++++++++++++++++++++++++++++
> > 3 files changed, 441 insertions(+)
> > create mode 100644 drivers/remoteproc/pru_rproc.c
> >
> > diff --git a/drivers/remoteproc/Kconfig b/drivers/remoteproc/Kconfig
> > index d99548fb5dde..3e3865a7cd78 100644
> > --- a/drivers/remoteproc/Kconfig
> > +++ b/drivers/remoteproc/Kconfig
> > @@ -125,6 +125,18 @@ config KEYSTONE_REMOTEPROC
> > It's safe to say N here if you're not interested in the Keystone
> > DSPs or just want to use a bare minimum kernel.
> >
> > +config PRU_REMOTEPROC
> > + tristate "TI PRU remoteproc support"
> > + depends on TI_PRUSS
> > + default TI_PRUSS
> > + help
> > + Support for TI PRU remote processors present within a PRU-ICSS
> > + subsystem via the remote processor framework.
> > +
> > + Say Y or M here to support the Programmable Realtime Unit (PRU)
> > + processors on various TI SoCs. It's safe to say N here if you're
> > + not interested in the PRU or if you are unsure.
> > +
> > config QCOM_PIL_INFO
> > tristate
> >
> > diff --git a/drivers/remoteproc/Makefile b/drivers/remoteproc/Makefile
> > index da2ace4ec86c..bb26c9e4ef9c 100644
> > --- a/drivers/remoteproc/Makefile
> > +++ b/drivers/remoteproc/Makefile
> > @@ -18,6 +18,7 @@ obj-$(CONFIG_OMAP_REMOTEPROC) += omap_remoteproc.o
> > obj-$(CONFIG_WKUP_M3_RPROC) += wkup_m3_rproc.o
> > obj-$(CONFIG_DA8XX_REMOTEPROC) += da8xx_remoteproc.o
> > obj-$(CONFIG_KEYSTONE_REMOTEPROC) += keystone_remoteproc.o
> > +obj-$(CONFIG_PRU_REMOTEPROC) += pru_rproc.o
> > obj-$(CONFIG_QCOM_PIL_INFO) += qcom_pil_info.o
> > obj-$(CONFIG_QCOM_RPROC_COMMON) += qcom_common.o
> > obj-$(CONFIG_QCOM_Q6V5_COMMON) += qcom_q6v5.o
> > diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> > new file mode 100644
> > index 000000000000..c94c8e965c21
> > --- /dev/null
> > +++ b/drivers/remoteproc/pru_rproc.c
> > @@ -0,0 +1,428 @@
> > +// SPDX-License-Identifier: GPL-2.0-only
> > +/*
> > + * PRU-ICSS remoteproc driver for various TI SoCs
> > + *
> > + * Copyright (C) 2014-2020 Texas Instruments Incorporated - https://www.ti.com/
> > + *
> > + * Author(s):
> > + * Suman Anna <[email protected]>
> > + * Andrew F. Davis <[email protected]>
> > + * Grzegorz Jaszczyk <[email protected]> for Texas Instruments
> > + */
> > +
> > +#include <linux/bitops.h>
> > +#include <linux/module.h>
> > +#include <linux/of_device.h>
> > +#include <linux/pruss_driver.h>
> > +#include <linux/remoteproc.h>
> > +
> > +#include "remoteproc_internal.h"
> > +#include "remoteproc_elf_helpers.h"
> > +
> > +/* PRU_ICSS_PRU_CTRL registers */
> > +#define PRU_CTRL_CTRL 0x0000
> > +#define PRU_CTRL_STS 0x0004
> > +
> > +/* CTRL register bit-fields */
> > +#define CTRL_CTRL_SOFT_RST_N BIT(0)
> > +#define CTRL_CTRL_EN BIT(1)
> > +#define CTRL_CTRL_SLEEPING BIT(2)
> > +#define CTRL_CTRL_CTR_EN BIT(3)
> > +#define CTRL_CTRL_SINGLE_STEP BIT(8)
> > +#define CTRL_CTRL_RUNSTATE BIT(15)
> > +
> > +/* PRU Core IRAM address masks */
> > +#define PRU0_IRAM_ADDR_MASK 0x34000
> > +#define PRU1_IRAM_ADDR_MASK 0x38000
> > +
> > +/* PRU device addresses for various type of PRU RAMs */
> > +#define PRU_IRAM_DA 0 /* Instruction RAM */
> > +#define PRU_PDRAM_DA 0 /* Primary Data RAM */
> > +#define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
> > +#define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
> > +
> > +/**
> > + * enum pru_iomem - PRU core memory/register range identifiers
> > + *
> > + * @PRU_IOMEM_IRAM: PRU Instruction RAM range
> > + * @PRU_IOMEM_CTRL: PRU Control register range
> > + * @PRU_IOMEM_DEBUG: PRU Debug register range
> > + * @PRU_IOMEM_MAX: just keep this one at the end
> > + */
> > +enum pru_iomem {
> > + PRU_IOMEM_IRAM = 0,
> > + PRU_IOMEM_CTRL,
> > + PRU_IOMEM_DEBUG,
> > + PRU_IOMEM_MAX,
> > +};
> > +
> > +/**
> > + * struct pru_rproc - PRU remoteproc structure
> > + * @id: id of the PRU core within the PRUSS
> > + * @dev: PRU core device pointer
> > + * @pruss: back-reference to parent PRUSS structure
> > + * @rproc: remoteproc pointer for this PRU core
> > + * @mem_regions: data for each of the PRU memory regions
> > + * @fw_name: name of firmware image used during loading
> > + */
> > +struct pru_rproc {
> > + int id;
> > + struct device *dev;
> > + struct pruss *pruss;
> > + struct rproc *rproc;
> > + struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> > + const char *fw_name;
> > +};
> > +
> > +static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
> > +{
> > + return readl_relaxed(pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> > +}
> > +
> > +static inline
> > +void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
> > +{
> > + writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> > +}
> > +
> > +static int pru_rproc_start(struct rproc *rproc)
> > +{
> > + struct device *dev = &rproc->dev;
> > + struct pru_rproc *pru = rproc->priv;
> > + u32 val;
> > +
> > + dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> > + pru->id, (rproc->bootaddr >> 2));
> > +
> > + val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
> > + pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> > +
> > + return 0;
> > +}
> > +
> > +static int pru_rproc_stop(struct rproc *rproc)
> > +{
> > + struct device *dev = &rproc->dev;
> > + struct pru_rproc *pru = rproc->priv;
> > + u32 val;
> > +
> > + dev_dbg(dev, "stopping PRU%d\n", pru->id);
> > +
> > + val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
> > + val &= ~CTRL_CTRL_EN;
> > + pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> > +
> > + return 0;
> > +}
> > +
> > +/*
> > + * Convert PRU device address (data spaces only) to kernel virtual address.
> > + *
> > + * Each PRU has access to all data memories within the PRUSS, accessible at
> > + * different ranges. So, look through both its primary and secondary Data
> > + * RAMs as well as any shared Data RAM to convert a PRU device address to
> > + * kernel virtual address. Data RAM0 is primary Data RAM for PRU0 and Data
> > + * RAM1 is primary Data RAM for PRU1.
> > + */
> > +static void *pru_d_da_to_va(struct pru_rproc *pru, u32 da, int len)
>
> The .da_to_va() ops argument has changed to using size_t in recent kernels, so
> let's convert the type from int to size_t for the len argument.
Good point. To be consistent I will do the same with the 'len'
argument for pru_da_to_va and pru_i_da_to_va.
>
> > +{
> > + struct pruss_mem_region dram0, dram1, shrd_ram;
> > + struct pruss *pruss = pru->pruss;
> > + u32 offset;
> > + void *va = NULL;
> > +
> > + if (len <= 0)
> > + return NULL;
>
> And we adjust this accordingly.
Sure.
>
> > +
> > + dram0 = pruss->mem_regions[PRUSS_MEM_DRAM0];
> > + dram1 = pruss->mem_regions[PRUSS_MEM_DRAM1];
> > + /* PRU1 has its local RAM addresses reversed */
> > + if (pru->id == 1)
> > + swap(dram0, dram1);
> > + shrd_ram = pruss->mem_regions[PRUSS_MEM_SHRD_RAM2];
> > +
> > + if (da >= PRU_PDRAM_DA && da + len <= PRU_PDRAM_DA + dram0.size) {
> > + offset = da - PRU_PDRAM_DA;
> > + va = (__force void *)(dram0.va + offset);
> > + } else if (da >= PRU_SDRAM_DA &&
> > + da + len <= PRU_SDRAM_DA + dram1.size) {
> > + offset = da - PRU_SDRAM_DA;
> > + va = (__force void *)(dram1.va + offset);
> > + } else if (da >= PRU_SHRDRAM_DA &&
> > + da + len <= PRU_SHRDRAM_DA + shrd_ram.size) {
> > + offset = da - PRU_SHRDRAM_DA;
> > + va = (__force void *)(shrd_ram.va + offset);
> > + }
> > +
> > + return va;
> > +}
> > +
> > +/*
> > + * Convert PRU device address (instruction space) to kernel virtual address.
> > + *
> > + * A PRU does not have an unified address space. Each PRU has its very own
> > + * private Instruction RAM, and its device address is identical to that of
> > + * its primary Data RAM device address.
> > + */
> > +static void *pru_i_da_to_va(struct pru_rproc *pru, u32 da, int len)
> > +{
> > + u32 offset;
> > + void *va = NULL;
> > +
> > + if (len <= 0)
> > + return NULL;
>
> Same comments as above.
Yes.
>
> > +
> > + if (da >= PRU_IRAM_DA &&
> > + da + len <= PRU_IRAM_DA + pru->mem_regions[PRU_IOMEM_IRAM].size) {
> > + offset = da - PRU_IRAM_DA;
> > + va = (__force void *)(pru->mem_regions[PRU_IOMEM_IRAM].va +
> > + offset);
> > + }
> > +
> > + return va;
> > +}
> > +
> > +/*
> > + * Provide address translations for only PRU Data RAMs through the remoteproc
> > + * core for any PRU client drivers. The PRU Instruction RAM access is restricted
> > + * only to the PRU loader code.
> > + */
> > +static void *pru_rproc_da_to_va(struct rproc *rproc, u64 da, size_t len)
> > +{
> > + struct pru_rproc *pru = rproc->priv;
> > +
> > + return pru_d_da_to_va(pru, da, len);
> > +}
> > +
> > +/* PRU-specific address translator used by PRU loader. */
> > +static void *pru_da_to_va(struct rproc *rproc, u64 da, int len, bool is_iram)
> > +{
> > + struct pru_rproc *pru = rproc->priv;
> > + void *va;
> > +
> > + if (is_iram)
> > + va = pru_i_da_to_va(pru, da, len);
> > + else
> > + va = pru_d_da_to_va(pru, da, len);
> > +
> > + return va;
> > +}
> > +
> > +static struct rproc_ops pru_rproc_ops = {
> > + .start = pru_rproc_start,
> > + .stop = pru_rproc_stop,
> > + .da_to_va = pru_rproc_da_to_va,
> > +};
> > +
> > +static int
> > +pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> > +{
> > + struct device *dev = &rproc->dev;
> > + struct elf32_hdr *ehdr;
> > + struct elf32_phdr *phdr;
> > + int i, ret = 0;
> > + const u8 *elf_data = fw->data;
> > +
> > + ehdr = (struct elf32_hdr *)elf_data;
> > + phdr = (struct elf32_phdr *)(elf_data + ehdr->e_phoff);
> > +
> > + /* go through the available ELF segments */
> > + for (i = 0; i < ehdr->e_phnum; i++, phdr++) {
> > + u32 da = phdr->p_paddr;
> > + u32 memsz = phdr->p_memsz;
> > + u32 filesz = phdr->p_filesz;
> > + u32 offset = phdr->p_offset;
> > + bool is_iram;
> > + void *ptr;
> > +
> > + if (phdr->p_type != PT_LOAD)
> > + continue;
> > +
> > + dev_dbg(dev, "phdr: type %d da 0x%x memsz 0x%x filesz 0x%x\n",
> > + phdr->p_type, da, memsz, filesz);
> > +
> > + if (filesz > memsz) {
> > + dev_err(dev, "bad phdr filesz 0x%x memsz 0x%x\n",
> > + filesz, memsz);
> > + ret = -EINVAL;
> > + break;
> > + }
> > +
> > + if (offset + filesz > fw->size) {
> > + dev_err(dev, "truncated fw: need 0x%x avail 0x%zx\n",
> > + offset + filesz, fw->size);
> > + ret = -EINVAL;
> > + break;
> > + }
> > +
> > + /* grab the kernel address for this device address */
> > + is_iram = phdr->p_flags & PF_X;
> > + ptr = pru_da_to_va(rproc, da, memsz, is_iram);
> > + if (!ptr) {
> > + dev_err(dev, "bad phdr da 0x%x mem 0x%x\n", da, memsz);
> > + ret = -EINVAL;
> > + break;
> > + }
> > +
> > + /* skip the memzero logic performed by remoteproc ELF loader */
> > + if (!phdr->p_filesz)
> > + continue;
> > +
> > + memcpy(ptr, elf_data + phdr->p_offset, filesz);
> > + }
> > +
> > + return ret;
> > +}
> > +
> > +static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
> > +{
> > + int ret;
> > +
> > + /* load optional rsc table */
> > + ret = rproc_elf_load_rsc_table(rproc, fw);
> > + if (ret == -EINVAL)
> > + dev_dbg(&rproc->dev, "no resource table found for this fw\n");
> > +
> > + return ret;
>
> This should return 0 if there is no resource table, otherwise this fails. You
> have this corrected in Patch 3, but please move that logic here.
Sure - sorry for that. I will move the correct logic from patch #3 to this one.
>
> > +}
> > +
> > +/*
> > + * Compute PRU id based on the IRAM addresses. The PRU IRAMs are
> > + * always at a particular offset within the PRUSS address space.
> > + */
> > +static int pru_rproc_set_id(struct pru_rproc *pru)
> > +{
> > + int ret = 0;
> > +
> > + switch (pru->mem_regions[PRU_IOMEM_IRAM].pa & 0x3ffff) {
>
> We can probably update this to use a macro for 0x3fffff.
Good idea, I will introduce PRU_IRAM_ADDR_MASK.
Thank you,
Grzegorz
>
> > + case PRU0_IRAM_ADDR_MASK:
> > + pru->id = 0;
> > + break;
> > + case PRU1_IRAM_ADDR_MASK:
> > + pru->id = 1;
> > + break;
> > + default:
> > + ret = -EINVAL;
> > + }
> > +
> > + return ret;
> > +}
> > +
> > +static int pru_rproc_probe(struct platform_device *pdev)
> > +{
> > + struct device *dev = &pdev->dev;
> > + struct device_node *np = dev->of_node;
> > + struct platform_device *ppdev = to_platform_device(dev->parent);
> > + struct pru_rproc *pru;
> > + const char *fw_name;
> > + struct rproc *rproc = NULL;
> > + struct resource *res;
> > + int i, ret;
> > + const char *mem_names[PRU_IOMEM_MAX] = { "iram", "control", "debug" };
> > +
> > + ret = of_property_read_string(np, "firmware-name", &fw_name);
> > + if (ret) {
> > + dev_err(dev, "unable to retrieve firmware-name %d\n", ret);
> > + return ret;
> > + }
> > +
> > + rproc = devm_rproc_alloc(dev, pdev->name, &pru_rproc_ops, fw_name,
> > + sizeof(*pru));
> > + if (!rproc) {
> > + dev_err(dev, "rproc_alloc failed\n");
> > + return -ENOMEM;
> > + }
> > + /* use a custom load function to deal with PRU-specific quirks */
> > + rproc->ops->load = pru_rproc_load_elf_segments;
> > +
> > + /* use a custom parse function to deal with PRU-specific resources */
> > + rproc->ops->parse_fw = pru_rproc_parse_fw;
> > +
> > + /* error recovery is not supported for PRUs */
> > + rproc->recovery_disabled = true;
> > +
> > + /*
> > + * rproc_add will auto-boot the processor normally, but this is not
> > + * desired with PRU client driven boot-flow methodology. A PRU
> > + * application/client driver will boot the corresponding PRU
> > + * remote-processor as part of its state machine either through the
> > + * remoteproc sysfs interface or through the equivalent kernel API.
> > + */
> > + rproc->auto_boot = false;
> > +
> > + pru = rproc->priv;
> > + pru->dev = dev;
> > + pru->pruss = platform_get_drvdata(ppdev);
> > + pru->rproc = rproc;
> > + pru->fw_name = fw_name;
> > +
> > + for (i = 0; i < ARRAY_SIZE(mem_names); i++) {
> > + res = platform_get_resource_byname(pdev, IORESOURCE_MEM,
> > + mem_names[i]);
> > + pru->mem_regions[i].va = devm_ioremap_resource(dev, res);
> > + if (IS_ERR(pru->mem_regions[i].va)) {
> > + dev_err(dev, "failed to parse and map memory resource %d %s\n",
> > + i, mem_names[i]);
> > + ret = PTR_ERR(pru->mem_regions[i].va);
> > + return ret;
> > + }
> > + pru->mem_regions[i].pa = res->start;
> > + pru->mem_regions[i].size = resource_size(res);
> > +
> > + dev_dbg(dev, "memory %8s: pa %pa size 0x%zx va %pK\n",
> > + mem_names[i], &pru->mem_regions[i].pa,
> > + pru->mem_regions[i].size, pru->mem_regions[i].va);
> > + }
> > +
> > + ret = pru_rproc_set_id(pru);
> > + if (ret < 0)
> > + return ret;
> > +
> > + platform_set_drvdata(pdev, rproc);
> > +
> > + ret = devm_rproc_add(dev, pru->rproc);
> > + if (ret) {
> > + dev_err(dev, "rproc_add failed: %d\n", ret);
> > + return ret;
> > + }
> > +
> > + dev_dbg(dev, "PRU rproc node %pOF probed successfully\n", np);
> > +
> > + return 0;
> > +}
> > +
> > +static int pru_rproc_remove(struct platform_device *pdev)
> > +{
> > + struct device *dev = &pdev->dev;
> > + struct rproc *rproc = platform_get_drvdata(pdev);
> > +
> > + dev_dbg(dev, "%s: removing rproc %s\n", __func__, rproc->name);
> > +
> > + return 0;
> > +}
> > +
> > +static const struct of_device_id pru_rproc_match[] = {
> > + { .compatible = "ti,am3356-pru", },
> > + { .compatible = "ti,am4376-pru", },
> > + { .compatible = "ti,am5728-pru", },
> > + { .compatible = "ti,k2g-pru", },
> > + {},
> > +};
> > +MODULE_DEVICE_TABLE(of, pru_rproc_match);
> > +
> > +static struct platform_driver pru_rproc_driver = {
> > + .driver = {
> > + .name = "pru-rproc",
> > + .of_match_table = pru_rproc_match,
> > + .suppress_bind_attrs = true,
> > + },
> > + .probe = pru_rproc_probe,
> > + .remove = pru_rproc_remove,
> > +};
> > +module_platform_driver(pru_rproc_driver);
> > +
> > +MODULE_AUTHOR("Suman Anna <[email protected]>");
> > +MODULE_AUTHOR("Andrew F. Davis <[email protected]>");
> > +MODULE_AUTHOR("Grzegorz Jaszczyk <[email protected]>");
> > +MODULE_DESCRIPTION("PRU-ICSS Remote Processor Driver");
> > +MODULE_LICENSE("GPL v2");
> >
>
Hi Suman,
On Tue, 17 Nov 2020 at 21:09, Suman Anna <[email protected]> wrote:
>
> Hi Greg,
>
> On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> > From: Suman Anna <[email protected]>
> >
> > The K3 AM65x family of SoCs have the next generation of the PRU-ICSS
> > processor subsystem, commonly referred to as ICSSG. Each ICSSG processor
> > subsystem on AM65x SR1.0 contains two primary PRU cores and two new
> > auxiliary PRU cores called RTUs. The AM65x SR2.0 SoCs have a revised
> > ICSSG IP that is based off the subsequent IP revision used on J721E
> > SoCs. This IP instance has two new custom auxiliary PRU cores called
> > Transmit PRUs (Tx_PRUs) in addition to the existing PRUs and RTUs.
> >
> > Each RTU and Tx_PRU cores have their own dedicated IRAM (smaller than
> > a PRU), Control and debug feature sets, but is different in terms of
> > sub-modules integrated around it and does not have the full capabilities
> > associated with a PRU core. The RTU core is typically used to aid a
> > PRU core in accelerating data transfers, while the Tx_PRU cores is
> > normally used to control the TX L2 FIFO if enabled in Ethernet
> > applications. Both can also be used to run independent applications.
> > The RTU and Tx_PRU cores though share the same Data RAMs as the PRU
> > cores, so the memories have to be partitioned carefully between different
> > applications. The new cores also support a new sub-module called Task
> > Manager to support two different context thread executions.
> >
> > Enhance the existing PRU remoteproc driver to support these new PRU, RTU
> > and Tx PRU cores by using specific compatibles. The initial names for the
> > firmware images for each PRU core are retrieved from DT nodes, and can
> > be adjusted through sysfs if required.
> >
> > The PRU remoteproc driver has to be specifically modified to use a
> > custom memcpy function within its ELF loader implementation for these
> > new cores in order to overcome a limitation with copying data into each
> > of the core's IRAM memories. These memory ports support only 4-byte
> > writes, and any sub-word order byte writes clear out the remaining
> > bytes other than the bytes being written within the containing word.
> > The default ARM64 memcpy also cannot be used as it throws an exception
> > when the preferred 8-byte copy operation is attempted. This choice is
> > made by using a state flag that is set only on K3 SoCs.
> >
> > Signed-off-by: Suman Anna <[email protected]>
> > Co-developed-by: Grzegorz Jaszczyk <[email protected]>
> > Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> > ---
> > drivers/remoteproc/pru_rproc.c | 141 ++++++++++++++++++++++++++++++---
> > 1 file changed, 132 insertions(+), 9 deletions(-)
> >
> > diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> > index 33806ddcbd5d..04c9f07799e2 100644
> > --- a/drivers/remoteproc/pru_rproc.c
> > +++ b/drivers/remoteproc/pru_rproc.c
> > @@ -46,9 +46,13 @@
> > #define PRU_DEBUG_GPREG(x) (0x0000 + (x) * 4)
> > #define PRU_DEBUG_CT_REG(x) (0x0080 + (x) * 4)
> >
> > -/* PRU Core IRAM address masks */
> > +/* PRU/RTU/Tx_PRU Core IRAM address masks */
> > #define PRU0_IRAM_ADDR_MASK 0x34000
> > #define PRU1_IRAM_ADDR_MASK 0x38000
> > +#define RTU0_IRAM_ADDR_MASK 0x4000
> > +#define RTU1_IRAM_ADDR_MASK 0x6000
> > +#define TX_PRU0_IRAM_ADDR_MASK 0xa000
> > +#define TX_PRU1_IRAM_ADDR_MASK 0xc000
> >
> > /* PRU device addresses for various type of PRU RAMs */
> > #define PRU_IRAM_DA 0 /* Instruction RAM */
> > @@ -73,12 +77,38 @@ enum pru_iomem {
> > PRU_IOMEM_MAX,
> > };
> >
> > +/**
> > + * enum pru_type - PRU core type identifier
> > + *
> > + * @PRU_TYPE_PRU: Programmable Real-time Unit
> > + * @PRU_TYPE_RTU: Auxiliary Programmable Real-Time Unit
> > + * @PRU_TYPE_TX_PRU: Transmit Programmable Real-Time Unit
> > + * @PRU_TYPE_MAX: just keep this one at the end
> > + */
> > +enum pru_type {
> > + PRU_TYPE_PRU = 0,
> > + PRU_TYPE_RTU,
> > + PRU_TYPE_TX_PRU,
> > + PRU_TYPE_MAX,
> > +};
> > +
> > +/**
> > + * struct pru_private_data - device data for a PRU core
> > + * @type: type of the PRU core (PRU, RTU, Tx_PRU)
> > + * @is_k3: flag used to identify the need for special load & event handling
> > + */
> > +struct pru_private_data {
> > + enum pru_type type;
> > + unsigned int is_k3 : 1;
> > +};
> > +
> > /**
> > * struct pru_rproc - PRU remoteproc structure
> > * @id: id of the PRU core within the PRUSS
> > * @dev: PRU core device pointer
> > * @pruss: back-reference to parent PRUSS structure
> > * @rproc: remoteproc pointer for this PRU core
> > + * @data: PRU core specific data
> > * @mem_regions: data for each of the PRU memory regions
> > * @fw_name: name of firmware image used during loading
> > * @mapped_irq: virtual interrupt numbers of created fw specific mapping
> > @@ -93,6 +123,7 @@ struct pru_rproc {
> > struct device *dev;
> > struct pruss *pruss;
> > struct rproc *rproc;
> > + const struct pru_private_data *data;
> > struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> > const char *fw_name;
> > int *mapped_irq;
> > @@ -318,11 +349,12 @@ static int pru_rproc_start(struct rproc *rproc)
> > {
> > struct device *dev = &rproc->dev;
> > struct pru_rproc *pru = rproc->priv;
> > + const char *names[PRU_TYPE_MAX] = { "PRU", "RTU", "Tx_PRU" };
> > u32 val;
> > int ret;
> >
> > - dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> > - pru->id, (rproc->bootaddr >> 2));
> > + dev_dbg(dev, "starting %s%d: entry-point = 0x%llx\n",
> > + names[pru->data->type], pru->id, (rproc->bootaddr >> 2));
> >
> > ret = pru_handle_intrmap(rproc);
> > /*
> > @@ -344,9 +376,10 @@ static int pru_rproc_stop(struct rproc *rproc)
> > {
> > struct device *dev = &rproc->dev;
> > struct pru_rproc *pru = rproc->priv;
> > + const char *names[PRU_TYPE_MAX] = { "PRU", "RTU", "Tx_PRU" };
> > u32 val;
> >
> > - dev_dbg(dev, "stopping PRU%d\n", pru->id);
> > + dev_dbg(dev, "stopping %s%d\n", names[pru->data->type], pru->id);
> >
> > val = pru_control_read_reg(pru, PRU_CTRL_CTRL);
> > val &= ~CTRL_CTRL_EN;
> > @@ -458,9 +491,53 @@ static struct rproc_ops pru_rproc_ops = {
> > .da_to_va = pru_rproc_da_to_va,
> > };
> >
> > +/*
> > + * Custom memory copy implementation for ICSSG PRU/RTU Cores
>
> Please update this to add Tx_PRU as well to the list here and in the below
> description.
Sure.
>
> > + *
> > + * The ICSSG PRU/RTU cores have a memory copying issue with IRAM memories, that
> > + * is not seen on previous generation SoCs. The data is reflected properly in
> > + * the IRAM memories only for integer (4-byte) copies. Any unaligned copies
> > + * result in all the other pre-existing bytes zeroed out within that 4-byte
> > + * boundary, thereby resulting in wrong text/code in the IRAMs. Also, the
> > + * IRAM memory port interface does not allow any 8-byte copies (as commonly
> > + * used by ARM64 memcpy implementation) and throws an exception. The DRAM
> > + * memory ports do not show this behavior. Use this custom copying function
> > + * to properly load the PRU/RTU firmware images on all memories for simplicity.
>
> This last line is obsolete now that we use regular memcpy for Data RAM copies.
Yes you are right. I will remove the last sentence.
Thank you,
Grzegorz
Hi Suman,
On Tue, 17 Nov 2020 at 21:40, Suman Anna <[email protected]> wrote:
>
> Hi Greg,
>
> On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> > The firmware blob can contain optional ELF sections: .resource_table
> > section and .pru_irq_map one. The second one contains the PRUSS
> > interrupt mapping description, which needs to be setup before powering
> > on the PRU core. To avoid RAM wastage this ELF section is not mapped to
> > any ELF segment (by the firmware linker) and therefore is not loaded to
> > PRU memory.
> >
> > The PRU interrupt configuration is handled within the PRUSS INTC irqchip
> > driver and leverages the system events to interrupt channels and host
> > interrupts mapping configuration. Relevant irq routing information is
> > passed through a special .pru_irq_map ELF section (for interrupts routed
> > to and used by PRU cores) or via the PRU application's device tree node
> > (for interrupts routed to and used by the main CPU). The mappings are
> > currently programmed during the booting/shutdown of the PRU.
> >
> > The interrupt configuration passed through .pru_irq_map ELF section is
> > optional. It varies on specific firmware functionality and therefore
> > have to be unwinded during PRU stop and performed again during
> > PRU start.
> >
> > Co-developed-by: Suman Anna <[email protected]>
> > Signed-off-by: Suman Anna <[email protected]>
> > Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> > ---
> > drivers/remoteproc/pru_rproc.c | 191 ++++++++++++++++++++++++++++++++-
> > drivers/remoteproc/pru_rproc.h | 46 ++++++++
> > 2 files changed, 236 insertions(+), 1 deletion(-)
> > create mode 100644 drivers/remoteproc/pru_rproc.h
> >
> > diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> > index c94c8e965c21..825e9c7e081b 100644
> > --- a/drivers/remoteproc/pru_rproc.c
> > +++ b/drivers/remoteproc/pru_rproc.c
> > @@ -11,13 +11,16 @@
> > */
> >
> > #include <linux/bitops.h>
> > +#include <linux/irqdomain.h>
> > #include <linux/module.h>
> > #include <linux/of_device.h>
> > +#include <linux/of_irq.h>
> > #include <linux/pruss_driver.h>
> > #include <linux/remoteproc.h>
> >
> > #include "remoteproc_internal.h"
> > #include "remoteproc_elf_helpers.h"
> > +#include "pru_rproc.h"
> >
> > /* PRU_ICSS_PRU_CTRL registers */
> > #define PRU_CTRL_CTRL 0x0000
> > @@ -41,6 +44,8 @@
> > #define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
> > #define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
> >
> > +#define MAX_PRU_SYS_EVENTS 160
> > +
> > /**
> > * enum pru_iomem - PRU core memory/register range identifiers
> > *
> > @@ -64,6 +69,10 @@ enum pru_iomem {
> > * @rproc: remoteproc pointer for this PRU core
> > * @mem_regions: data for each of the PRU memory regions
> > * @fw_name: name of firmware image used during loading
> > + * @mapped_irq: virtual interrupt numbers of created fw specific mapping
> > + * @pru_interrupt_map: pointer to interrupt mapping description (firmware)
> > + * @pru_interrupt_map_sz: pru_interrupt_map size
> > + * @evt_count: number of mapped events
> > */
> > struct pru_rproc {
> > int id;
> > @@ -72,6 +81,10 @@ struct pru_rproc {
> > struct rproc *rproc;
> > struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> > const char *fw_name;
> > + int *mapped_irq;
> > + struct pru_irq_rsc *pru_interrupt_map;
> > + size_t pru_interrupt_map_sz;
> > + ssize_t evt_count;
>
> Do you really need this to be ssize_t type?
You are right, it is not needed. I will use size_t type instead and
modify relevant while loop.
>
> > };
> >
> > static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
> > @@ -85,15 +98,107 @@ void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
> > writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> > }
> >
> > +static void pru_dispose_irq_mapping(struct pru_rproc *pru)
> > +{
> > + while (--pru->evt_count >= 0) {
> > + if (pru->mapped_irq[pru->evt_count] > 0)
> > + irq_dispose_mapping(pru->mapped_irq[pru->evt_count]);
> > + }
> > +
> > + kfree(pru->mapped_irq);
> > +}
> > +
> > +/*
> > + * Parse the custom PRU interrupt map resource and configure the INTC
> > + * appropriately.
> > + */
> > +static int pru_handle_intrmap(struct rproc *rproc)
> > +{
> > + struct device *dev = rproc->dev.parent;
> > + struct pru_rproc *pru = rproc->priv;
> > + struct pru_irq_rsc *rsc = pru->pru_interrupt_map;
> > + struct irq_fwspec fwspec;
> > + struct device_node *irq_parent;
> > + int i, ret = 0;
> > +
> > + /* not having pru_interrupt_map is not an error */
> > + if (!rsc)
> > + return 0;
> > +
> > + /* currently supporting only type 0 */
> > + if (rsc->type != 0) {
> > + dev_err(dev, "unsupported rsc type: %d\n", rsc->type);
> > + return -EINVAL;
> > + }
> > +
> > + if (rsc->num_evts < 0 || rsc->num_evts > MAX_PRU_SYS_EVENTS)
> > + return -EINVAL;
> > +
> > + if (sizeof(*rsc) + rsc->num_evts * sizeof(struct pruss_int_map) !=
> > + pru->pru_interrupt_map_sz)
> > + return -EINVAL;
> > +
> > + pru->evt_count = rsc->num_evts;
> > + pru->mapped_irq = kcalloc(pru->evt_count, sizeof(int), GFP_KERNEL);
> > + if (!pru->mapped_irq)
> > + return -ENOMEM;
> > +
> > + /*
> > + * parse and fill in system event to interrupt channel and
> > + * channel-to-host mapping
> > + */
> > + irq_parent = of_irq_find_parent(pru->dev->of_node);
> > + if (!irq_parent) {
> > + kfree(pru->mapped_irq);
> > + return -ENODEV;
> > + }
> > +
> > + fwspec.fwnode = of_node_to_fwnode(irq_parent);
> > + fwspec.param_count = 3;
> > + for (i = 0; i < pru->evt_count; i++) {
> > + fwspec.param[0] = rsc->pru_intc_map[i].event;
> > + fwspec.param[1] = rsc->pru_intc_map[i].chnl;
> > + fwspec.param[2] = rsc->pru_intc_map[i].host;
> > +
> > + dev_dbg(dev, "mapping%d: event %d, chnl %d, host %d\n",
> > + i, fwspec.param[0], fwspec.param[1], fwspec.param[2]);
> > +
> > + pru->mapped_irq[i] = irq_create_fwspec_mapping(&fwspec);
> > + if (pru->mapped_irq[i] < 0) {
> > + dev_err(dev, "failed to get virq\n");
> > + ret = pru->mapped_irq[i];
> > + goto map_fail;
> > + }
> > + }
> > +
> > + return ret;
> > +
> > +map_fail:
> > + pru_dispose_irq_mapping(pru);
> > +
> > + return ret;
> > +}
> > +
> > static int pru_rproc_start(struct rproc *rproc)
> > {
> > struct device *dev = &rproc->dev;
> > struct pru_rproc *pru = rproc->priv;
> > u32 val;
> > + int ret;
> >
> > dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> > pru->id, (rproc->bootaddr >> 2));
> >
> > + ret = pru_handle_intrmap(rproc);
> > + /*
> > + * reset references to pru interrupt map - they will stop being valid
> > + * after rproc_start returns
> > + */
> > + pru->pru_interrupt_map = NULL;
> > + pru->pru_interrupt_map_sz = 0;
> > + if (ret)
> > + return ret;
> > +
> > val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
> > pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> >
> > @@ -112,6 +217,10 @@ static int pru_rproc_stop(struct rproc *rproc)
> > val &= ~CTRL_CTRL_EN;
> > pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> >
> > + /* dispose irq mapping - new firmware can provide new mapping */
> > + if (pru->mapped_irq)
> > + pru_dispose_irq_mapping(pru);
> > +
> > return 0;
> > }
> >
> > @@ -274,16 +383,96 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> > return ret;
> > }
> >
> > +static const void *
> > +pru_rproc_find_interrupt_map(struct device *dev, const struct firmware *fw)
> > +{
> > + const void *shdr, *name_table_shdr;
> > + const char *name_table;
> > + const u8 *elf_data = (void *)fw->data;
> > + u8 class = fw_elf_get_class(fw);
> > + size_t fw_size = fw->size;
> > + const void *ehdr = elf_data;
> > + u16 shnum = elf_hdr_get_e_shnum(class, ehdr);
> > + u32 elf_shdr_get_size = elf_size_of_shdr(class);
> > + u16 shstrndx = elf_hdr_get_e_shstrndx(class, ehdr);
> > + int i;
> > +
> > + /* first, get the section header according to the elf class */
> > + shdr = elf_data + elf_hdr_get_e_shoff(class, ehdr);
> > + /* compute name table section header entry in shdr array */
> > + name_table_shdr = shdr + (shstrndx * elf_shdr_get_size);
> > + /* finally, compute the name table section address in elf */
> > + name_table = elf_data + elf_shdr_get_sh_offset(class, name_table_shdr);
>
> I see you used the style influenced by the remoteproc_elf_loader code. PRUs are
> all 32-bit, so we need not use this strictly. I am ok with this style, but
> prefer consistent usage style between this function and
> pru_rproc_load_elf_segments().
Ok. I will get rid of generic elf helpers macros usage and will stick
with elf32_* related structs instead. This will make it similar to
pru_rproc_load_elf_segments() in terms of style.
>
> > +
> > + for (i = 0; i < shnum; i++, shdr += elf_shdr_get_size) {
> > + u64 size = elf_shdr_get_sh_size(class, shdr);
> > + u64 offset = elf_shdr_get_sh_offset(class, shdr);
> > + u32 name = elf_shdr_get_sh_name(class, shdr);
> > +
> > + if (strcmp(name_table + name, ".pru_irq_map"))
> > + continue;
> > +
> > + /* make sure we have the entire table */
> > + if (offset + size > fw_size || offset + size < size) {
> > + dev_err(dev, "interrupt map sec truncated\n");
>
> sec can confuse developers, suggest rephrasing this trace, something like
> ".pru_irq_map section truncated"
Ok. I will also update the comment to: "make sure we have the entire irq map"
>
> > + return ERR_PTR(-EINVAL);
> > + }
> > +
> > + /* make sure table has at least the header */
> > + if (sizeof(struct pru_irq_rsc) > size) {
> > + dev_err(dev, "header-less interrupt map sec\n");
>
> same comment as above
Sure, I will use "header-less .pru_irq_map section\n" and update the
comment above to: "make sure irq map has at least the header"
>
> > + return ERR_PTR(-EINVAL);
> > + }
> > +
> > + return shdr;
> > + }
> > +
> > + dev_dbg(dev, "no .pru_irq_map section found for this fw\n");
> > +
> > + return NULL;
> > +}
> > +
> > +/*
> > + * Usa a custom parse_fw callback function for dealing with standard
> > + * resource table and a PRU-specific custom ELF section.
> > + *
> > + * The firmware blob can contain optional ELF sections: .resource_table section
> > + * and .pru_irq_map one. The second one contains the PRUSS interrupt mapping
> > + * description, which needs to be setup before powering on the PRU core. To
> > + * avoid RAM wastage this ELF section is not mapped to any ELF segment (by the
> > + * firmware linker) and therefore is not loaded to PRU memory.
>
> Some of this description can move to Patch 2 as well.
I thought about it before posting but IMO this function without
.pru_irq_map handling is kind of self explaining. Especially when
taking into account comment and deb_dbg message from the function body
of previous patch. Nevertheless I can add:
/*
* Usa a custom parse_fw callback function for dealing with optional resource
* table.
*/
to patch #2 and then overwrite it in patch #3, but not sure if it is
the best option.
>
> > + */
> > static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
> > {
> > + struct device *dev = &rproc->dev;
> > + struct pru_rproc *pru = rproc->priv;
> > + const u8 *elf_data = fw->data;
> > + const void *shdr;
> > + u8 class = fw_elf_get_class(fw);
> > + u64 sh_offset;
> > int ret;
> >
> > /* load optional rsc table */
> > ret = rproc_elf_load_rsc_table(rproc, fw);
> > if (ret == -EINVAL)
> > dev_dbg(&rproc->dev, "no resource table found for this fw\n");
> > + else if (ret)
> > + return ret;
>
> This hunk should be part of Patch 2.
You are right.
>
> >
> > - return ret;
> > + /* find .pru_interrupt_map section, not having it is not an error */
> > + shdr = pru_rproc_find_interrupt_map(dev, fw);
> > + if (IS_ERR(shdr))
> > + return PTR_ERR(shdr);
> > +
> > + if (!shdr)
> > + return 0;
> > +
> > + /* preserve pointer to PRU interrupt map together with it size */
> > + sh_offset = elf_shdr_get_sh_offset(class, shdr);
> > + pru->pru_interrupt_map = (struct pru_irq_rsc *)(elf_data + sh_offset);
> > + pru->pru_interrupt_map_sz = elf_shdr_get_sh_size(class, shdr);
> > +
> > + return 0;
>
> And this one as well.
Correct,
Thank you for your review,
Grzegorz
Hi Greg,
On 11/18/20 9:27 AM, Grzegorz Jaszczyk wrote:
> Hi Suman,
>
> On Tue, 17 Nov 2020 at 21:40, Suman Anna <[email protected]> wrote:
>>
>> Hi Greg,
>>
>> On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
>>> The firmware blob can contain optional ELF sections: .resource_table
>>> section and .pru_irq_map one. The second one contains the PRUSS
>>> interrupt mapping description, which needs to be setup before powering
>>> on the PRU core. To avoid RAM wastage this ELF section is not mapped to
>>> any ELF segment (by the firmware linker) and therefore is not loaded to
>>> PRU memory.
>>>
>>> The PRU interrupt configuration is handled within the PRUSS INTC irqchip
>>> driver and leverages the system events to interrupt channels and host
>>> interrupts mapping configuration. Relevant irq routing information is
>>> passed through a special .pru_irq_map ELF section (for interrupts routed
>>> to and used by PRU cores) or via the PRU application's device tree node
>>> (for interrupts routed to and used by the main CPU). The mappings are
>>> currently programmed during the booting/shutdown of the PRU.
>>>
>>> The interrupt configuration passed through .pru_irq_map ELF section is
>>> optional. It varies on specific firmware functionality and therefore
>>> have to be unwinded during PRU stop and performed again during
>>> PRU start.
>>>
>>> Co-developed-by: Suman Anna <[email protected]>
>>> Signed-off-by: Suman Anna <[email protected]>
>>> Signed-off-by: Grzegorz Jaszczyk <[email protected]>
>>> ---
>>> drivers/remoteproc/pru_rproc.c | 191 ++++++++++++++++++++++++++++++++-
>>> drivers/remoteproc/pru_rproc.h | 46 ++++++++
>>> 2 files changed, 236 insertions(+), 1 deletion(-)
>>> create mode 100644 drivers/remoteproc/pru_rproc.h
>>>
>>> diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
>>> index c94c8e965c21..825e9c7e081b 100644
>>> --- a/drivers/remoteproc/pru_rproc.c
>>> +++ b/drivers/remoteproc/pru_rproc.c
>>> @@ -11,13 +11,16 @@
>>> */
>>>
>>> #include <linux/bitops.h>
>>> +#include <linux/irqdomain.h>
>>> #include <linux/module.h>
>>> #include <linux/of_device.h>
>>> +#include <linux/of_irq.h>
>>> #include <linux/pruss_driver.h>
>>> #include <linux/remoteproc.h>
>>>
>>> #include "remoteproc_internal.h"
>>> #include "remoteproc_elf_helpers.h"
>>> +#include "pru_rproc.h"
>>>
>>> /* PRU_ICSS_PRU_CTRL registers */
>>> #define PRU_CTRL_CTRL 0x0000
>>> @@ -41,6 +44,8 @@
>>> #define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
>>> #define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
>>>
>>> +#define MAX_PRU_SYS_EVENTS 160
>>> +
>>> /**
>>> * enum pru_iomem - PRU core memory/register range identifiers
>>> *
>>> @@ -64,6 +69,10 @@ enum pru_iomem {
>>> * @rproc: remoteproc pointer for this PRU core
>>> * @mem_regions: data for each of the PRU memory regions
>>> * @fw_name: name of firmware image used during loading
>>> + * @mapped_irq: virtual interrupt numbers of created fw specific mapping
>>> + * @pru_interrupt_map: pointer to interrupt mapping description (firmware)
>>> + * @pru_interrupt_map_sz: pru_interrupt_map size
>>> + * @evt_count: number of mapped events
>>> */
>>> struct pru_rproc {
>>> int id;
>>> @@ -72,6 +81,10 @@ struct pru_rproc {
>>> struct rproc *rproc;
>>> struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
>>> const char *fw_name;
>>> + int *mapped_irq;
>>> + struct pru_irq_rsc *pru_interrupt_map;
>>> + size_t pru_interrupt_map_sz;
>>> + ssize_t evt_count;
>>
>> Do you really need this to be ssize_t type?
>
> You are right, it is not needed. I will use size_t type instead and
> modify relevant while loop.
Hmm, size_t is ok, but perhaps the same type you used struct pru_irq_rsc is better.
>
>>
>>> };
>>>
>>> static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
>>> @@ -85,15 +98,107 @@ void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
>>> writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
>>> }
>>>
>>> +static void pru_dispose_irq_mapping(struct pru_rproc *pru)
>>> +{
>>> + while (--pru->evt_count >= 0) {
>>> + if (pru->mapped_irq[pru->evt_count] > 0)
>>> + irq_dispose_mapping(pru->mapped_irq[pru->evt_count]);
>>> + }
>>> +
>>> + kfree(pru->mapped_irq);
>>> +}
>>> +
>>> +/*
>>> + * Parse the custom PRU interrupt map resource and configure the INTC
>>> + * appropriately.
>>> + */
>>> +static int pru_handle_intrmap(struct rproc *rproc)
>>> +{
>>> + struct device *dev = rproc->dev.parent;
>>> + struct pru_rproc *pru = rproc->priv;
>>> + struct pru_irq_rsc *rsc = pru->pru_interrupt_map;
>>> + struct irq_fwspec fwspec;
>>> + struct device_node *irq_parent;
>>> + int i, ret = 0;
>>> +
>>> + /* not having pru_interrupt_map is not an error */
>>> + if (!rsc)
>>> + return 0;
>>> +
>>> + /* currently supporting only type 0 */
>>> + if (rsc->type != 0) {
>>> + dev_err(dev, "unsupported rsc type: %d\n", rsc->type);
>>> + return -EINVAL;
>>> + }
>>> +
>>> + if (rsc->num_evts < 0 || rsc->num_evts > MAX_PRU_SYS_EVENTS)
>>> + return -EINVAL;
>>> +
>>> + if (sizeof(*rsc) + rsc->num_evts * sizeof(struct pruss_int_map) !=
>>> + pru->pru_interrupt_map_sz)
>>> + return -EINVAL;
>>> +
>>> + pru->evt_count = rsc->num_evts;
>>> + pru->mapped_irq = kcalloc(pru->evt_count, sizeof(int), GFP_KERNEL);
>>> + if (!pru->mapped_irq)
>>> + return -ENOMEM;
>>> +
>>> + /*
>>> + * parse and fill in system event to interrupt channel and
>>> + * channel-to-host mapping
>>> + */
>>> + irq_parent = of_irq_find_parent(pru->dev->of_node);
>>> + if (!irq_parent) {
>>> + kfree(pru->mapped_irq);
>>> + return -ENODEV;
>>> + }
>>> +
>>> + fwspec.fwnode = of_node_to_fwnode(irq_parent);
>>> + fwspec.param_count = 3;
>>> + for (i = 0; i < pru->evt_count; i++) {
>>> + fwspec.param[0] = rsc->pru_intc_map[i].event;
>>> + fwspec.param[1] = rsc->pru_intc_map[i].chnl;
>>> + fwspec.param[2] = rsc->pru_intc_map[i].host;
>>> +
>>> + dev_dbg(dev, "mapping%d: event %d, chnl %d, host %d\n",
>>> + i, fwspec.param[0], fwspec.param[1], fwspec.param[2]);
>>> +
>>> + pru->mapped_irq[i] = irq_create_fwspec_mapping(&fwspec);
>>> + if (pru->mapped_irq[i] < 0) {
>>> + dev_err(dev, "failed to get virq\n");
>>> + ret = pru->mapped_irq[i];
>>> + goto map_fail;
>>> + }
>>> + }
>>> +
>>> + return ret;
>>> +
>>> +map_fail:
>>> + pru_dispose_irq_mapping(pru);
>>> +
>>> + return ret;
>>> +}
>>> +
>>> static int pru_rproc_start(struct rproc *rproc)
>>> {
>>> struct device *dev = &rproc->dev;
>>> struct pru_rproc *pru = rproc->priv;
>>> u32 val;
>>> + int ret;
>>>
>>> dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
>>> pru->id, (rproc->bootaddr >> 2));
>>>
>>> + ret = pru_handle_intrmap(rproc);
>>> + /*
>>> + * reset references to pru interrupt map - they will stop being valid
>>> + * after rproc_start returns
>>> + */
>>> + pru->pru_interrupt_map = NULL;
>>> + pru->pru_interrupt_map_sz = 0;
>>> + if (ret)
>>> + return ret;
>>> +
>>> val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
>>> pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
>>>
>>> @@ -112,6 +217,10 @@ static int pru_rproc_stop(struct rproc *rproc)
>>> val &= ~CTRL_CTRL_EN;
>>> pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
>>>
>>> + /* dispose irq mapping - new firmware can provide new mapping */
>>> + if (pru->mapped_irq)
>>> + pru_dispose_irq_mapping(pru);
>>> +
>>> return 0;
>>> }
>>>
>>> @@ -274,16 +383,96 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
>>> return ret;
>>> }
>>>
>>> +static const void *
>>> +pru_rproc_find_interrupt_map(struct device *dev, const struct firmware *fw)
>>> +{
>>> + const void *shdr, *name_table_shdr;
>>> + const char *name_table;
>>> + const u8 *elf_data = (void *)fw->data;
>>> + u8 class = fw_elf_get_class(fw);
>>> + size_t fw_size = fw->size;
>>> + const void *ehdr = elf_data;
>>> + u16 shnum = elf_hdr_get_e_shnum(class, ehdr);
>>> + u32 elf_shdr_get_size = elf_size_of_shdr(class);
>>> + u16 shstrndx = elf_hdr_get_e_shstrndx(class, ehdr);
>>> + int i;
>>> +
>>> + /* first, get the section header according to the elf class */
>>> + shdr = elf_data + elf_hdr_get_e_shoff(class, ehdr);
>>> + /* compute name table section header entry in shdr array */
>>> + name_table_shdr = shdr + (shstrndx * elf_shdr_get_size);
>>> + /* finally, compute the name table section address in elf */
>>> + name_table = elf_data + elf_shdr_get_sh_offset(class, name_table_shdr);
>>
>> I see you used the style influenced by the remoteproc_elf_loader code. PRUs are
>> all 32-bit, so we need not use this strictly. I am ok with this style, but
>> prefer consistent usage style between this function and
>> pru_rproc_load_elf_segments().
>
> Ok. I will get rid of generic elf helpers macros usage and will stick
> with elf32_* related structs instead. This will make it similar to
> pru_rproc_load_elf_segments() in terms of style.
Yeah, ok with me.
>
>>
>>> +
>>> + for (i = 0; i < shnum; i++, shdr += elf_shdr_get_size) {
>>> + u64 size = elf_shdr_get_sh_size(class, shdr);
>>> + u64 offset = elf_shdr_get_sh_offset(class, shdr);
>>> + u32 name = elf_shdr_get_sh_name(class, shdr);
>>> +
>>> + if (strcmp(name_table + name, ".pru_irq_map"))
>>> + continue;
>>> +
>>> + /* make sure we have the entire table */
>>> + if (offset + size > fw_size || offset + size < size) {
>>> + dev_err(dev, "interrupt map sec truncated\n");
>>
>> sec can confuse developers, suggest rephrasing this trace, something like
>> ".pru_irq_map section truncated"
>
> Ok. I will also update the comment to: "make sure we have the entire irq map"
Please don't use "we" in debug traces, and just the summary of the error trace.
>
>>
>>> + return ERR_PTR(-EINVAL);
>>> + }
>>> +
>>> + /* make sure table has at least the header */
>>> + if (sizeof(struct pru_irq_rsc) > size) {
>>> + dev_err(dev, "header-less interrupt map sec\n");
>>
>> same comment as above
>
> Sure, I will use "header-less .pru_irq_map section\n" and update the
> comment above to: "make sure irq map has at least the header"
similar to above
>
>>
>>> + return ERR_PTR(-EINVAL);
>>> + }
>>> +
>>> + return shdr;
>>> + }
>>> +
>>> + dev_dbg(dev, "no .pru_irq_map section found for this fw\n");
>>> +
>>> + return NULL;
>>> +}
>>> +
>>> +/*
>>> + * Usa a custom parse_fw callback function for dealing with standard
>>> + * resource table and a PRU-specific custom ELF section.
>>> + *
>>> + * The firmware blob can contain optional ELF sections: .resource_table section
>>> + * and .pru_irq_map one. The second one contains the PRUSS interrupt mapping
>>> + * description, which needs to be setup before powering on the PRU core. To
>>> + * avoid RAM wastage this ELF section is not mapped to any ELF segment (by the
>>> + * firmware linker) and therefore is not loaded to PRU memory.
>>
>> Some of this description can move to Patch 2 as well.
>
> I thought about it before posting but IMO this function without
> .pru_irq_map handling is kind of self explaining. Especially when
> taking into account comment and deb_dbg message from the function body
> of previous patch. Nevertheless I can add:
> /*
> * Usa a custom parse_fw callback function for dealing with optional resource
> * table.
> */
Or you can keep the first two lines generic, "... dealing with PRU firmware
specific sections", and only add the details in the second paragraph in this
patch. The commit in code above resource table + these two links should suffice
for patch 2.
>
> to patch #2 and then overwrite it in patch #3, but not sure if it is
> the best option.
>
>>
>>> + */
>>> static int pru_rproc_parse_fw(struct rproc *rproc, const struct firmware *fw)
>>> {
>>> + struct device *dev = &rproc->dev;
>>> + struct pru_rproc *pru = rproc->priv;
>>> + const u8 *elf_data = fw->data;
>>> + const void *shdr;
>>> + u8 class = fw_elf_get_class(fw);
>>> + u64 sh_offset;
>>> int ret;
>>>
>>> /* load optional rsc table */
>>> ret = rproc_elf_load_rsc_table(rproc, fw);
>>> if (ret == -EINVAL)
>>> dev_dbg(&rproc->dev, "no resource table found for this fw\n");
>>> + else if (ret)
>>> + return ret;
>>
>> This hunk should be part of Patch 2.
>
> You are right.
>
>>
>>>
>>> - return ret;
>>> + /* find .pru_interrupt_map section, not having it is not an error */
>>> + shdr = pru_rproc_find_interrupt_map(dev, fw);
>>> + if (IS_ERR(shdr))
>>> + return PTR_ERR(shdr);
>>> +
>>> + if (!shdr)
>>> + return 0;
>>> +
>>> + /* preserve pointer to PRU interrupt map together with it size */
>>> + sh_offset = elf_shdr_get_sh_offset(class, shdr);
>>> + pru->pru_interrupt_map = (struct pru_irq_rsc *)(elf_data + sh_offset);
>>> + pru->pru_interrupt_map_sz = elf_shdr_get_sh_size(class, shdr);
>>> +
>>> + return 0;
>>
>> And this one as well.
>
> Correct,
>
> Thank you for your review,
> Grzegorz
regards
Suman
Hi Suman,
On Thu, 19 Nov 2020 at 00:04, Suman Anna <[email protected]> wrote:
>
> Hi Greg,
>
> On 11/18/20 9:27 AM, Grzegorz Jaszczyk wrote:
> > Hi Suman,
> >
> > On Tue, 17 Nov 2020 at 21:40, Suman Anna <[email protected]> wrote:
> >>
> >> Hi Greg,
> >>
> >> On 11/14/20 2:46 AM, Grzegorz Jaszczyk wrote:
> >>> The firmware blob can contain optional ELF sections: .resource_table
> >>> section and .pru_irq_map one. The second one contains the PRUSS
> >>> interrupt mapping description, which needs to be setup before powering
> >>> on the PRU core. To avoid RAM wastage this ELF section is not mapped to
> >>> any ELF segment (by the firmware linker) and therefore is not loaded to
> >>> PRU memory.
> >>>
> >>> The PRU interrupt configuration is handled within the PRUSS INTC irqchip
> >>> driver and leverages the system events to interrupt channels and host
> >>> interrupts mapping configuration. Relevant irq routing information is
> >>> passed through a special .pru_irq_map ELF section (for interrupts routed
> >>> to and used by PRU cores) or via the PRU application's device tree node
> >>> (for interrupts routed to and used by the main CPU). The mappings are
> >>> currently programmed during the booting/shutdown of the PRU.
> >>>
> >>> The interrupt configuration passed through .pru_irq_map ELF section is
> >>> optional. It varies on specific firmware functionality and therefore
> >>> have to be unwinded during PRU stop and performed again during
> >>> PRU start.
> >>>
> >>> Co-developed-by: Suman Anna <[email protected]>
> >>> Signed-off-by: Suman Anna <[email protected]>
> >>> Signed-off-by: Grzegorz Jaszczyk <[email protected]>
> >>> ---
> >>> drivers/remoteproc/pru_rproc.c | 191 ++++++++++++++++++++++++++++++++-
> >>> drivers/remoteproc/pru_rproc.h | 46 ++++++++
> >>> 2 files changed, 236 insertions(+), 1 deletion(-)
> >>> create mode 100644 drivers/remoteproc/pru_rproc.h
> >>>
> >>> diff --git a/drivers/remoteproc/pru_rproc.c b/drivers/remoteproc/pru_rproc.c
> >>> index c94c8e965c21..825e9c7e081b 100644
> >>> --- a/drivers/remoteproc/pru_rproc.c
> >>> +++ b/drivers/remoteproc/pru_rproc.c
> >>> @@ -11,13 +11,16 @@
> >>> */
> >>>
> >>> #include <linux/bitops.h>
> >>> +#include <linux/irqdomain.h>
> >>> #include <linux/module.h>
> >>> #include <linux/of_device.h>
> >>> +#include <linux/of_irq.h>
> >>> #include <linux/pruss_driver.h>
> >>> #include <linux/remoteproc.h>
> >>>
> >>> #include "remoteproc_internal.h"
> >>> #include "remoteproc_elf_helpers.h"
> >>> +#include "pru_rproc.h"
> >>>
> >>> /* PRU_ICSS_PRU_CTRL registers */
> >>> #define PRU_CTRL_CTRL 0x0000
> >>> @@ -41,6 +44,8 @@
> >>> #define PRU_SDRAM_DA 0x2000 /* Secondary Data RAM */
> >>> #define PRU_SHRDRAM_DA 0x10000 /* Shared Data RAM */
> >>>
> >>> +#define MAX_PRU_SYS_EVENTS 160
> >>> +
> >>> /**
> >>> * enum pru_iomem - PRU core memory/register range identifiers
> >>> *
> >>> @@ -64,6 +69,10 @@ enum pru_iomem {
> >>> * @rproc: remoteproc pointer for this PRU core
> >>> * @mem_regions: data for each of the PRU memory regions
> >>> * @fw_name: name of firmware image used during loading
> >>> + * @mapped_irq: virtual interrupt numbers of created fw specific mapping
> >>> + * @pru_interrupt_map: pointer to interrupt mapping description (firmware)
> >>> + * @pru_interrupt_map_sz: pru_interrupt_map size
> >>> + * @evt_count: number of mapped events
> >>> */
> >>> struct pru_rproc {
> >>> int id;
> >>> @@ -72,6 +81,10 @@ struct pru_rproc {
> >>> struct rproc *rproc;
> >>> struct pruss_mem_region mem_regions[PRU_IOMEM_MAX];
> >>> const char *fw_name;
> >>> + int *mapped_irq;
> >>> + struct pru_irq_rsc *pru_interrupt_map;
> >>> + size_t pru_interrupt_map_sz;
> >>> + ssize_t evt_count;
> >>
> >> Do you really need this to be ssize_t type?
> >
> > You are right, it is not needed. I will use size_t type instead and
> > modify relevant while loop.
>
> Hmm, size_t is ok, but perhaps the same type you used struct pru_irq_rsc is better.
Ok.
>
> >
> >>
> >>> };
> >>>
> >>> static inline u32 pru_control_read_reg(struct pru_rproc *pru, unsigned int reg)
> >>> @@ -85,15 +98,107 @@ void pru_control_write_reg(struct pru_rproc *pru, unsigned int reg, u32 val)
> >>> writel_relaxed(val, pru->mem_regions[PRU_IOMEM_CTRL].va + reg);
> >>> }
> >>>
> >>> +static void pru_dispose_irq_mapping(struct pru_rproc *pru)
> >>> +{
> >>> + while (--pru->evt_count >= 0) {
> >>> + if (pru->mapped_irq[pru->evt_count] > 0)
> >>> + irq_dispose_mapping(pru->mapped_irq[pru->evt_count]);
> >>> + }
> >>> +
> >>> + kfree(pru->mapped_irq);
> >>> +}
> >>> +
> >>> +/*
> >>> + * Parse the custom PRU interrupt map resource and configure the INTC
> >>> + * appropriately.
> >>> + */
> >>> +static int pru_handle_intrmap(struct rproc *rproc)
> >>> +{
> >>> + struct device *dev = rproc->dev.parent;
> >>> + struct pru_rproc *pru = rproc->priv;
> >>> + struct pru_irq_rsc *rsc = pru->pru_interrupt_map;
> >>> + struct irq_fwspec fwspec;
> >>> + struct device_node *irq_parent;
> >>> + int i, ret = 0;
> >>> +
> >>> + /* not having pru_interrupt_map is not an error */
> >>> + if (!rsc)
> >>> + return 0;
> >>> +
> >>> + /* currently supporting only type 0 */
> >>> + if (rsc->type != 0) {
> >>> + dev_err(dev, "unsupported rsc type: %d\n", rsc->type);
> >>> + return -EINVAL;
> >>> + }
> >>> +
> >>> + if (rsc->num_evts < 0 || rsc->num_evts > MAX_PRU_SYS_EVENTS)
> >>> + return -EINVAL;
> >>> +
> >>> + if (sizeof(*rsc) + rsc->num_evts * sizeof(struct pruss_int_map) !=
> >>> + pru->pru_interrupt_map_sz)
> >>> + return -EINVAL;
> >>> +
> >>> + pru->evt_count = rsc->num_evts;
> >>> + pru->mapped_irq = kcalloc(pru->evt_count, sizeof(int), GFP_KERNEL);
> >>> + if (!pru->mapped_irq)
> >>> + return -ENOMEM;
> >>> +
> >>> + /*
> >>> + * parse and fill in system event to interrupt channel and
> >>> + * channel-to-host mapping
> >>> + */
> >>> + irq_parent = of_irq_find_parent(pru->dev->of_node);
> >>> + if (!irq_parent) {
> >>> + kfree(pru->mapped_irq);
> >>> + return -ENODEV;
> >>> + }
> >>> +
> >>> + fwspec.fwnode = of_node_to_fwnode(irq_parent);
> >>> + fwspec.param_count = 3;
> >>> + for (i = 0; i < pru->evt_count; i++) {
> >>> + fwspec.param[0] = rsc->pru_intc_map[i].event;
> >>> + fwspec.param[1] = rsc->pru_intc_map[i].chnl;
> >>> + fwspec.param[2] = rsc->pru_intc_map[i].host;
> >>> +
> >>> + dev_dbg(dev, "mapping%d: event %d, chnl %d, host %d\n",
> >>> + i, fwspec.param[0], fwspec.param[1], fwspec.param[2]);
> >>> +
> >>> + pru->mapped_irq[i] = irq_create_fwspec_mapping(&fwspec);
> >>> + if (pru->mapped_irq[i] < 0) {
> >>> + dev_err(dev, "failed to get virq\n");
> >>> + ret = pru->mapped_irq[i];
> >>> + goto map_fail;
> >>> + }
> >>> + }
> >>> +
> >>> + return ret;
> >>> +
> >>> +map_fail:
> >>> + pru_dispose_irq_mapping(pru);
> >>> +
> >>> + return ret;
> >>> +}
> >>> +
> >>> static int pru_rproc_start(struct rproc *rproc)
> >>> {
> >>> struct device *dev = &rproc->dev;
> >>> struct pru_rproc *pru = rproc->priv;
> >>> u32 val;
> >>> + int ret;
> >>>
> >>> dev_dbg(dev, "starting PRU%d: entry-point = 0x%llx\n",
> >>> pru->id, (rproc->bootaddr >> 2));
> >>>
> >>> + ret = pru_handle_intrmap(rproc);
> >>> + /*
> >>> + * reset references to pru interrupt map - they will stop being valid
> >>> + * after rproc_start returns
> >>> + */
> >>> + pru->pru_interrupt_map = NULL;
> >>> + pru->pru_interrupt_map_sz = 0;
> >>> + if (ret)
> >>> + return ret;
> >>> +
> >>> val = CTRL_CTRL_EN | ((rproc->bootaddr >> 2) << 16);
> >>> pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> >>>
> >>> @@ -112,6 +217,10 @@ static int pru_rproc_stop(struct rproc *rproc)
> >>> val &= ~CTRL_CTRL_EN;
> >>> pru_control_write_reg(pru, PRU_CTRL_CTRL, val);
> >>>
> >>> + /* dispose irq mapping - new firmware can provide new mapping */
> >>> + if (pru->mapped_irq)
> >>> + pru_dispose_irq_mapping(pru);
> >>> +
> >>> return 0;
> >>> }
> >>>
> >>> @@ -274,16 +383,96 @@ pru_rproc_load_elf_segments(struct rproc *rproc, const struct firmware *fw)
> >>> return ret;
> >>> }
> >>>
> >>> +static const void *
> >>> +pru_rproc_find_interrupt_map(struct device *dev, const struct firmware *fw)
> >>> +{
> >>> + const void *shdr, *name_table_shdr;
> >>> + const char *name_table;
> >>> + const u8 *elf_data = (void *)fw->data;
> >>> + u8 class = fw_elf_get_class(fw);
> >>> + size_t fw_size = fw->size;
> >>> + const void *ehdr = elf_data;
> >>> + u16 shnum = elf_hdr_get_e_shnum(class, ehdr);
> >>> + u32 elf_shdr_get_size = elf_size_of_shdr(class);
> >>> + u16 shstrndx = elf_hdr_get_e_shstrndx(class, ehdr);
> >>> + int i;
> >>> +
> >>> + /* first, get the section header according to the elf class */
> >>> + shdr = elf_data + elf_hdr_get_e_shoff(class, ehdr);
> >>> + /* compute name table section header entry in shdr array */
> >>> + name_table_shdr = shdr + (shstrndx * elf_shdr_get_size);
> >>> + /* finally, compute the name table section address in elf */
> >>> + name_table = elf_data + elf_shdr_get_sh_offset(class, name_table_shdr);
> >>
> >> I see you used the style influenced by the remoteproc_elf_loader code. PRUs are
> >> all 32-bit, so we need not use this strictly. I am ok with this style, but
> >> prefer consistent usage style between this function and
> >> pru_rproc_load_elf_segments().
> >
> > Ok. I will get rid of generic elf helpers macros usage and will stick
> > with elf32_* related structs instead. This will make it similar to
> > pru_rproc_load_elf_segments() in terms of style.
>
> Yeah, ok with me.
>
> >
> >>
> >>> +
> >>> + for (i = 0; i < shnum; i++, shdr += elf_shdr_get_size) {
> >>> + u64 size = elf_shdr_get_sh_size(class, shdr);
> >>> + u64 offset = elf_shdr_get_sh_offset(class, shdr);
> >>> + u32 name = elf_shdr_get_sh_name(class, shdr);
> >>> +
> >>> + if (strcmp(name_table + name, ".pru_irq_map"))
> >>> + continue;
> >>> +
> >>> + /* make sure we have the entire table */
> >>> + if (offset + size > fw_size || offset + size < size) {
> >>> + dev_err(dev, "interrupt map sec truncated\n");
> >>
> >> sec can confuse developers, suggest rephrasing this trace, something like
> >> ".pru_irq_map section truncated"
> >
> > Ok. I will also update the comment to: "make sure we have the entire irq map"
>
> Please don't use "we" in debug traces, and just the summary of the error trace.
I think you've misunderstood my answer: I agreed with your suggestion
to rephrase this trace to ".pru_irq_map section truncated" but
additionally I wanted to say that I will also update the comment (not
the dev_err trace) which is located two lines above.
>
> >
> >>
> >>> + return ERR_PTR(-EINVAL);
> >>> + }
> >>> +
> >>> + /* make sure table has at least the header */
> >>> + if (sizeof(struct pru_irq_rsc) > size) {
> >>> + dev_err(dev, "header-less interrupt map sec\n");
> >>
> >> same comment as above
> >
> > Sure, I will use "header-less .pru_irq_map section\n" and update the
> > comment above to: "make sure irq map has at least the header"
>
> similar to above
The same as above :)
>
> >
> >>
> >>> + return ERR_PTR(-EINVAL);
> >>> + }
> >>> +
> >>> + return shdr;
> >>> + }
> >>> +
> >>> + dev_dbg(dev, "no .pru_irq_map section found for this fw\n");
> >>> +
> >>> + return NULL;
> >>> +}
> >>> +
> >>> +/*
> >>> + * Usa a custom parse_fw callback function for dealing with standard
> >>> + * resource table and a PRU-specific custom ELF section.
> >>> + *
> >>> + * The firmware blob can contain optional ELF sections: .resource_table section
> >>> + * and .pru_irq_map one. The second one contains the PRUSS interrupt mapping
> >>> + * description, which needs to be setup before powering on the PRU core. To
> >>> + * avoid RAM wastage this ELF section is not mapped to any ELF segment (by the
> >>> + * firmware linker) and therefore is not loaded to PRU memory.
> >>
> >> Some of this description can move to Patch 2 as well.
> >
> > I thought about it before posting but IMO this function without
> > .pru_irq_map handling is kind of self explaining. Especially when
> > taking into account comment and deb_dbg message from the function body
> > of previous patch. Nevertheless I can add:
> > /*
> > * Usa a custom parse_fw callback function for dealing with optional resource
> > * table.
> > */
>
> Or you can keep the first two lines generic, "... dealing with PRU firmware
> specific sections", and only add the details in the second paragraph in this
> patch. The commit in code above resource table + these two links should suffice
> for patch 2.
Ok.
Thank you,
Grzegorz