From: Xiaofei Tan <tanxiaofei@huawei.com> driver inclusion category: feature bugzilla: https://gitee.com/openeuler/kernel/issues/IAO2AR CVE: NA -------------------------------- Kunpeng SoC support several memory repair capability. Normally, such repairs are done in firmware startup stage or triggered by BMC(Baseboard Management Controller) and done in firmware in runtime stage. OS can't perceive such capability. Some support online repair, while others can't. One reason of not supporting online repair is that the memory is in use and the hardware does not support backpressure on the primary access channel. In order to support more online memory repair, we seek solutions from OS. That is try to isolate the memory page first, and do repair, and de-isloate after repair done. This patch support online ACLS(Adapter Cache Line Sparing) repair, and sPPR(Soft Post Package Repair). All need firmware support first. Signed-off-by: Xiaofei Tan <tanxiaofei@huawei.com> Signed-off-by: Junhao He <hejunhao3@huawei.com> --- MAINTAINERS | 6 + drivers/soc/hisilicon/Kconfig | 14 + drivers/soc/hisilicon/Makefile | 1 + drivers/soc/hisilicon/hisi_mem_ras.c | 582 +++++++++++++++++++++++++++ drivers/soc/hisilicon/hisi_mem_ras.h | 112 ++++++ 5 files changed, 715 insertions(+) create mode 100644 drivers/soc/hisilicon/hisi_mem_ras.c create mode 100644 drivers/soc/hisilicon/hisi_mem_ras.h diff --git a/MAINTAINERS b/MAINTAINERS index 83b9596bbe7c..e0d9f78c75d1 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7952,6 +7952,12 @@ F: Documentation/ABI/testing/sysfs-devices-platform-kunpeng_hccs F: drivers/soc/hisilicon/kunpeng_hccs.c F: drivers/soc/hisilicon/kunpeng_hccs.h +HISILICON KUNPENG SOC MEMORY RAS DRIVER +M: Xiaofei Tan <tanxiaofei@huawei.com> +S: Maintained +F: drivers/soc/hisilicon/hisi_mem_ras.c +F: drivers/soc/hisilicon/hisi_mem_ras.h + HISILICON LPC BUS DRIVER M: john.garry@huawei.com S: Maintained diff --git a/drivers/soc/hisilicon/Kconfig b/drivers/soc/hisilicon/Kconfig index ec5be0b59f95..0e79de44a01f 100644 --- a/drivers/soc/hisilicon/Kconfig +++ b/drivers/soc/hisilicon/Kconfig @@ -58,4 +58,18 @@ config HISI_L0 for user to alloc and lock memory If not sure say no. + +config HISI_MEM_RAS + tristate "Add RAS support for Hisilicon memory device" + depends on ACPI + depends on PCC + depends on ARM64 || COMPILE_TEST + help + Add RAS driver for Hisilicon memory device to repair hardware error of + memory devices in OS side. The driver interacts with firmware through PCC + and provides sysfs files as user-mode interfaces.It could support both + HBM and DDR memory. + + If not sure say no. + endmenu diff --git a/drivers/soc/hisilicon/Makefile b/drivers/soc/hisilicon/Makefile index d91b8ed30306..aac8f8463e34 100644 --- a/drivers/soc/hisilicon/Makefile +++ b/drivers/soc/hisilicon/Makefile @@ -7,3 +7,4 @@ obj-$(CONFIG_ARM64_PBHA) += pbha.o hisi_l3t-objs := hisi_lockdown.o l3t.o obj-$(CONFIG_HISI_L3T) += hisi_l3t.o obj-$(CONFIG_HISI_L0) += hisi_l0.o +obj-$(CONFIG_HISI_MEM_RAS) += hisi_mem_ras.o diff --git a/drivers/soc/hisilicon/hisi_mem_ras.c b/drivers/soc/hisilicon/hisi_mem_ras.c new file mode 100644 index 000000000000..765132cf1c04 --- /dev/null +++ b/drivers/soc/hisilicon/hisi_mem_ras.c @@ -0,0 +1,582 @@ +// SPDX-License-Identifier: GPL-2.0 +/* + * Copyright (C) Huawei Technologies Co., Ltd. 2024. All rights reserved. + */ + +#include <linux/acpi.h> +#include <linux/module.h> +#include <linux/mm.h> +#include <linux/platform_device.h> +#include <acpi/pcc.h> +#include "hisi_mem_ras.h" + +#define DRV_NAME "hisi_mem_ras" +#define MAX_PCC_CMD_RETRIES 500ULL + +struct hisi_mem_register_ctx { + struct device *dev; + u8 chan_id; + int err; +}; + +static acpi_status hisi_mem_get_chan_id_cb(struct acpi_resource *ares, void *context) +{ + struct acpi_resource_generic_register *reg; + struct hisi_mem_register_ctx *ctx = context; + + if (ares->type != ACPI_RESOURCE_TYPE_GENERIC_REGISTER) + return AE_OK; + + reg = &ares->data.generic_reg; + if (reg->space_id != ACPI_ADR_SPACE_PLATFORM_COMM) { + dev_err(ctx->dev, "Bad register resource.\n"); + ctx->err = -EINVAL; + return AE_ERROR; + } + ctx->chan_id = reg->access_size; + + return AE_OK; +} + +static int hisi_mem_get_pcc_chan_id(struct hisi_mem_dev *hdev) +{ + struct platform_device *pdev = hdev->pdev; + struct acpi_device *adev = ACPI_COMPANION(&pdev->dev); + struct hisi_mem_register_ctx ctx = {0}; + acpi_handle handle = adev->handle; + acpi_status status; + + if (!acpi_has_method(handle, METHOD_NAME__CRS)) { + dev_err(&pdev->dev, "No _CRS method.\n"); + return -ENODEV; + } + + ctx.dev = &pdev->dev; + status = acpi_walk_resources(handle, METHOD_NAME__CRS, + hisi_mem_get_chan_id_cb, &ctx); + if (ACPI_FAILURE(status)) + return ctx.err; + + hdev->chan_id = ctx.chan_id; + return 0; +} + +static void hisi_mem_chan_tx_done(struct mbox_client *cl, void *mssg, int ret) +{ + if (ret < 0) + pr_debug("TX did not complete: CMD sent:0x%x, ret:%d\n", + *(u8 *)mssg, ret); + else + pr_debug("TX completed. CMD sent:0x%x, ret:%d\n", + *(u8 *)mssg, ret); +} + +static void hisi_mem_pcc_rx_callback(struct mbox_client *cl, void *mssg) +{ + struct hisi_mem_mbox_client_info *cl_info = + container_of(cl, struct hisi_mem_mbox_client_info, client); + + complete(&cl_info->done); +} + +static void hisi_mem_unregister_pcc_channel(struct hisi_mem_dev *hdev) +{ + struct hisi_mem_mbox_client_info *cl_info = &hdev->cl_info; + + pcc_mbox_free_channel(cl_info->pcc_chan); +} + +static int hisi_mem_register_pcc_channel(struct hisi_mem_dev *hdev) +{ + struct hisi_mem_mbox_client_info *cl_info = &hdev->cl_info; + struct mbox_client *cl = &cl_info->client; + struct platform_device *pdev = hdev->pdev; + struct pcc_mbox_chan *pcc_chan; + int rc; + + cl->dev = &pdev->dev; + cl->tx_block = false; + cl->knows_txdone = true; + cl->tx_done = hisi_mem_chan_tx_done; + cl->rx_callback = hisi_mem_pcc_rx_callback; + pcc_chan = pcc_mbox_request_channel(cl, hdev->chan_id); + if (IS_ERR(pcc_chan)) + return dev_err_probe(&pdev->dev, -ENODEV, "PCC channel request failed.\n"); + + if (pcc_chan->shmem_size > HISI_MEM_PCC_SHARE_MEM_BYTES_MAX || + pcc_chan->shmem_size < HISI_MEM_PCC_SHARE_MEM_BYTES_MIN) { + rc = dev_err_probe(&pdev->dev, -EINVAL, "Unsupported PCC shmem size 0x%llx.\n", + pcc_chan->shmem_size); + goto err_mbx_channel_free; + } + + /* + * pcc_chan->latency is just a nominal value. In reality the remote + * processor could be much slower to reply. So add an arbitrary amount + * of wait on top of nominal. + */ + cl_info->deadline_us = MAX_PCC_CMD_RETRIES * pcc_chan->latency; + cl_info->pcc_chan = pcc_chan; + init_completion(&cl_info->done); + if (!pcc_chan->mchan->mbox->txdone_irq) { + rc = dev_err_probe(&pdev->dev, -EINVAL, "PCC IRQ in PCCT isn't supported.\n"); + goto err_mbx_channel_free; + } + + cl_info->pcc_comm_addr = devm_ioremap(&pdev->dev, pcc_chan->shmem_base_addr, + pcc_chan->shmem_size); + if (!cl_info->pcc_comm_addr) { + rc = dev_err_probe(&pdev->dev, -ENOMEM, + "Failed to ioremap PCC communication region for channel-%u.\n", + hdev->chan_id); + goto err_mbx_channel_free; + } + + return 0; + +err_mbx_channel_free: + pcc_mbox_free_channel(cl_info->pcc_chan); + return rc; +} + +static int hisi_mem_wait_cmd_complete_by_irq(struct hisi_mem_dev *hdev) +{ + struct hisi_mem_mbox_client_info *cl_info = &hdev->cl_info; + struct platform_device *pdev = hdev->pdev; + + if (!wait_for_completion_timeout(&cl_info->done, usecs_to_jiffies(cl_info->deadline_us))) { + dev_err(&pdev->dev, "PCC command executed timeout!\n"); + return -ETIMEDOUT; + } + + return 0; +} + +static inline void hisi_mem_fill_ext_pcc_shared_mem_region(struct hisi_mem_dev *hdev, u8 cmd, + struct hisi_mem_desc *desc, + void __iomem *comm_space, + u32 space_size) +{ + struct hisi_mem_mbox_client_info *cl_info = &hdev->cl_info; + struct acpi_pcct_ext_pcc_shared_memory tmp = { + .signature = PCC_SIGNATURE | hdev->chan_id, + .flags = PCC_CMD_COMPLETION_NOTIFY, + .length = cl_info->pcc_chan->shmem_size, + .command = cmd, + }; + + memcpy_toio(hdev->cl_info.pcc_comm_addr, &tmp, sizeof(tmp)); + + /* Copy the message to the PCC comm space */ + memcpy_toio(comm_space, desc, space_size); +} + +static int hisi_mem_pcc_cmd_send(struct hisi_mem_dev *hdev, u8 cmd, + struct hisi_mem_desc *desc) +{ + struct hisi_mem_mbox_client_info *cl_info = &hdev->cl_info; + struct platform_device *pdev = hdev->pdev; + void __iomem *comm_space; + u32 space_size; + int ret; + + mutex_lock(&hdev->lock); + comm_space = cl_info->pcc_comm_addr + sizeof(struct acpi_pcct_ext_pcc_shared_memory); + space_size = cl_info->pcc_chan->shmem_size - sizeof(struct acpi_pcct_ext_pcc_shared_memory); + hisi_mem_fill_ext_pcc_shared_mem_region(hdev, cmd, desc, comm_space, space_size); + reinit_completion(&cl_info->done); + + /* Ring doorbell */ + ret = mbox_send_message(cl_info->pcc_chan->mchan, &cmd); + if (ret < 0) { + dev_err(&pdev->dev, "Send PCC mbox message failed, ret = %d.\n", ret); + goto end; + } + + ret = hisi_mem_wait_cmd_complete_by_irq(hdev); + if (ret) + goto end; + + /* Copy response data */ + memcpy_fromio(desc, comm_space, space_size); + +end: + mbox_chan_txdone(cl_info->pcc_chan->mchan, ret); + mutex_unlock(&hdev->lock); + return ret; +} + +static int check_fw_ret_status(struct device *dev, u8 status, char *cmd_desc) +{ + if (status == HISI_MEM_RAS_NO_RES) { + dev_info(dev, "%s failed due to no resource.\n", cmd_desc); + return -ENOSPC; + } else if (status != HISI_MEM_RAS_OK) { + dev_info(dev, "%s failed due to error code %u.\n", cmd_desc, status); + return -ENXIO; + } + + return 0; +} + +static struct page *trans_and_check_paddr(struct device *dev, const char *buf, u64 *paddr) +{ + struct page *page; + u64 paddr_tmp; + + if (kstrtoull(buf, 16, &paddr_tmp)) + return NULL; + + page = pfn_to_online_page(__phys_to_pfn(paddr_tmp)); + if (!page) { + dev_info(dev, "The page of paddr 0x%llx is not online page.\n", paddr_tmp); + return NULL; + } + + *paddr = paddr_tmp; + return page; +} + +static ssize_t acls_query_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + struct platform_device *pdev = container_of(dev, struct platform_device, dev); + struct hisi_mem_dev *hdev = platform_get_drvdata(pdev); + struct hisi_mem_desc desc = {0}; + struct page *page; + u64 paddr; + int ret; + + page = trans_and_check_paddr(dev, buf, &paddr); + if (!page) + return -EINVAL; + + desc.req.req_head.subcmd = HISI_MEM_RAS_QUERY; + desc.req.data[0] = lower_32_bits(paddr); + desc.req.data[1] = upper_32_bits(paddr); + ret = hisi_mem_pcc_cmd_send(hdev, HISI_MEM_RAS_ACLS, &desc); + if (ret) { + dev_err(dev, "ACLS query failed, ret = %d.\n", ret); + return -EIO; + } + + ret = check_fw_ret_status(dev, desc.rsp.rsp_head.ret_status, "ACLS query"); + if (ret) + return ret; + + return count; +} +static DEVICE_ATTR_WO(acls_query); + +static ssize_t acls_repair_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + struct platform_device *pdev = container_of(dev, struct platform_device, dev); + struct hisi_mem_dev *hdev = platform_get_drvdata(pdev); + struct hisi_mem_desc desc = {0}; + struct page *page; + u64 paddr; + int ret; + + page = trans_and_check_paddr(dev, buf, &paddr); + if (!page) + return -EINVAL; + + if (hdev->mem_type == MEMORY_TYPE_HBM && !PageHWPoison(page)) { + dev_info(dev, "The page of paddr 0x%llx is not unpoisoned.\n", paddr); + return -EIO; + } + + desc.req.req_head.subcmd = HISI_MEM_RAS_DO_REPAIR; + desc.req.data[0] = lower_32_bits(paddr); + desc.req.data[1] = upper_32_bits(paddr); + ret = hisi_mem_pcc_cmd_send(hdev, HISI_MEM_RAS_ACLS, &desc); + if (ret) { + dev_err(dev, "ACLS repair failed, ret = %d.\n", ret); + return -EIO; + } + + ret = check_fw_ret_status(dev, desc.rsp.rsp_head.ret_status, "ACLS repair"); + if (ret) + return ret; + + return count; +} +static DEVICE_ATTR_WO(acls_repair); + +static ssize_t sppr_query_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + struct platform_device *pdev = container_of(dev, struct platform_device, dev); + struct hisi_mem_dev *hdev = platform_get_drvdata(pdev); + struct hisi_mem_desc desc = {0}; + struct page *page; + u64 paddr; + int ret; + + page = trans_and_check_paddr(dev, buf, &paddr); + if (!page) + return -EINVAL; + + desc.req.req_head.subcmd = HISI_MEM_RAS_QUERY; + desc.req.data[0] = lower_32_bits(paddr); + desc.req.data[1] = upper_32_bits(paddr); + ret = hisi_mem_pcc_cmd_send(hdev, HISI_MEM_RAS_SPPR, &desc); + if (ret) { + dev_err(dev, "SPPR query failed, ret = %d.\n", ret); + return -EIO; + } + + ret = check_fw_ret_status(dev, desc.rsp.rsp_head.ret_status, "SPPR query"); + if (ret) + return ret; + + return count; +} +static DEVICE_ATTR_WO(sppr_query); + +static ssize_t sppr_repair_store(struct device *dev, struct device_attribute *attr, + const char *buf, size_t count) +{ + struct platform_device *pdev = container_of(dev, struct platform_device, dev); + struct hisi_mem_dev *hdev = platform_get_drvdata(pdev); + struct hisi_mem_desc desc = {0}; + struct page *page; + u64 paddr; + int ret; + + page = trans_and_check_paddr(dev, buf, &paddr); + if (!page) + return -EINVAL; + + if (hdev->mem_type == MEMORY_TYPE_HBM && !PageHWPoison(page)) { + dev_info(dev, "The page of paddr 0x%llx is not poisoned.\n", paddr); + return -EIO; + } + + desc.req.req_head.subcmd = HISI_MEM_RAS_DO_REPAIR; + desc.req.data[0] = lower_32_bits(paddr); + desc.req.data[1] = upper_32_bits(paddr); + ret = hisi_mem_pcc_cmd_send(hdev, HISI_MEM_RAS_SPPR, &desc); + if (ret) { + dev_err(dev, "SPPR repair failed, ret = %d.\n", ret); + return -EIO; + } + + ret = check_fw_ret_status(dev, desc.rsp.rsp_head.ret_status, "SPPR repair"); + if (ret) + return ret; + + return count; +} +static DEVICE_ATTR_WO(sppr_repair); + +static const char * const memory_type_name[MEMORY_TYPE_MAX] = { + "HBM", + "DDR", +}; + +static ssize_t memory_type_show(struct device *dev, struct device_attribute *attr, + char *buf) +{ + struct platform_device *pdev = container_of(dev, struct platform_device, dev); + struct hisi_mem_dev *hdev = platform_get_drvdata(pdev); + int ret; + + if (hdev->mem_type < MEMORY_TYPE_MAX) + ret = sysfs_emit(buf, "%s\n", memory_type_name[hdev->mem_type]); + else + ret = sysfs_emit(buf, "Unknown\n"); + + return ret; +} +static DEVICE_ATTR_RO(memory_type); + +static ssize_t memory_type_supported_show(struct device *dev, struct device_attribute *attr, + char *buf) +{ + int ret = 0; + int i; + + ret += sysfs_emit_at(buf, ret, "%s", memory_type_name[0]); + for (i = 1; i < MEMORY_TYPE_MAX; i++) + ret += sysfs_emit_at(buf, ret, " %s", memory_type_name[i]); + + ret += sysfs_emit_at(buf, ret, "\n"); + + return ret; +} +static DEVICE_ATTR_RO(memory_type_supported); + +static int hisi_mem_ras_get_attribute(struct hisi_mem_dev *hdev) +{ + struct platform_device *pdev = hdev->pdev; + struct hisi_mem_desc desc = {0}; + int ret; + + ret = hisi_mem_pcc_cmd_send(hdev, HISI_MEM_RAS_CAP, &desc); + if (ret) + return dev_err_probe(&pdev->dev, -EIO, "Get attribute failed, ret = %d.\n", ret); + + if (desc.rsp.rsp_head.ret_status) + return dev_err_probe(&pdev->dev, -EIO, "Device in bad status %u.\n", + desc.rsp.rsp_head.ret_status); + + if (desc.rsp.rsp_head.mem_type >= MEMORY_TYPE_MAX) + return dev_err_probe(&pdev->dev, -EIO, "Unsupported memory type %u.\n", + desc.rsp.rsp_head.mem_type); + + hdev->mem_type = desc.rsp.rsp_head.mem_type; + hdev->ras_cap = desc.rsp.rsp_head.ras_cap; + + return 0; +} + +static struct attribute *memory_type_attrs[] = { + &dev_attr_memory_type.attr, + &dev_attr_memory_type_supported.attr, + NULL, +}; + +static struct attribute_group memory_type_attr_group = { + .attrs = memory_type_attrs, +}; + +static struct attribute *acls_attrs[] = { + &dev_attr_acls_query.attr, + &dev_attr_acls_repair.attr, + NULL, +}; + +static struct attribute_group acls_attr_group = { + .attrs = acls_attrs, +}; + +static struct attribute *sppr_attrs[] = { + &dev_attr_sppr_query.attr, + &dev_attr_sppr_repair.attr, + NULL, +}; + +static struct attribute_group sppr_attr_group = { + .attrs = sppr_attrs, +}; + +static int hisi_mem_create_sysfs_files(struct hisi_mem_dev *hdev) +{ + struct platform_device *pdev = hdev->pdev; + int ret; + + if (hdev->ras_cap & HISI_MEM_CAP_ACLS_EN) { + ret = sysfs_create_group(&pdev->dev.kobj, &acls_attr_group); + if (ret) { + dev_err(&pdev->dev, "Create ACLS sysfs group failed, ret = %d.\n", ret); + return ret; + } + } + + if (hdev->ras_cap & HISI_MEM_CAP_SPPR_EN) { + ret = sysfs_create_group(&pdev->dev.kobj, &sppr_attr_group); + if (ret) { + dev_err(&pdev->dev, "Create SPPR sysfs group failed, ret = %d.\n", ret); + goto acls_files; + } + } + + ret = sysfs_create_group(&pdev->dev.kobj, &memory_type_attr_group); + if (ret) { + dev_err(&pdev->dev, "Create memory type file failed, ret =%d.\n", ret); + goto sppr_files; + } + + return 0; + +sppr_files: + if (hdev->ras_cap & HISI_MEM_CAP_SPPR_EN) + sysfs_remove_group(&pdev->dev.kobj, &sppr_attr_group); + +acls_files: + if (hdev->ras_cap & HISI_MEM_CAP_ACLS_EN) + sysfs_remove_group(&pdev->dev.kobj, &acls_attr_group); + + return ret; +} + +static void hisi_mem_remove_sysfs_files(struct hisi_mem_dev *hdev) +{ + struct platform_device *pdev = hdev->pdev; + + sysfs_remove_group(&pdev->dev.kobj, &memory_type_attr_group); + if (hdev->ras_cap & HISI_MEM_CAP_SPPR_EN) + sysfs_remove_group(&pdev->dev.kobj, &sppr_attr_group); + + if (hdev->ras_cap & HISI_MEM_CAP_ACLS_EN) + sysfs_remove_group(&pdev->dev.kobj, &acls_attr_group); +} + +static int hisi_mem_ras_probe(struct platform_device *pdev) +{ + struct hisi_mem_dev *hdev; + int ret; + + hdev = devm_kzalloc(&pdev->dev, sizeof(struct hisi_mem_dev), GFP_KERNEL); + if (!hdev) + return -ENOMEM; + + hdev->pdev = pdev; + mutex_init(&hdev->lock); + platform_set_drvdata(pdev, hdev); + ret = hisi_mem_get_pcc_chan_id(hdev); + if (ret) + return ret; + + ret = hisi_mem_register_pcc_channel(hdev); + if (ret) + return ret; + + ret = hisi_mem_ras_get_attribute(hdev); + if (ret) + goto unregister_pcc_chan; + + ret = hisi_mem_create_sysfs_files(hdev); + if (ret) + goto unregister_pcc_chan; + + return 0; + +unregister_pcc_chan: + hisi_mem_unregister_pcc_channel(hdev); + + return ret; +} + +static int hisi_mem_ras_remove(struct platform_device *pdev) +{ + struct hisi_mem_dev *hdev = platform_get_drvdata(pdev); + + hisi_mem_remove_sysfs_files(hdev); + hisi_mem_unregister_pcc_channel(hdev); + + return 0; +} + +static const struct acpi_device_id hisi_mem_acpi_match[] = { + { "HISI0521", 0 }, + { } +}; +MODULE_DEVICE_TABLE(acpi, hisi_mem_acpi_match); + +static struct platform_driver hisi_mem_ras_driver = { + .probe = hisi_mem_ras_probe, + .remove = hisi_mem_ras_remove, + .driver = { + .name = "hisi_mem_ras", + .acpi_match_table = hisi_mem_acpi_match, + }, +}; +module_platform_driver(hisi_mem_ras_driver); + +MODULE_LICENSE("GPL"); +MODULE_AUTHOR("Xiaofei Tan <tanxiaofei@huawei.com>"); +MODULE_DESCRIPTION("HISILICON Memory RAS driver"); +MODULE_ALIAS("platform:" DRV_NAME); diff --git a/drivers/soc/hisilicon/hisi_mem_ras.h b/drivers/soc/hisilicon/hisi_mem_ras.h new file mode 100644 index 000000000000..9ab9561b434d --- /dev/null +++ b/drivers/soc/hisilicon/hisi_mem_ras.h @@ -0,0 +1,112 @@ +/* SPDX-License-Identifier: GPL-2.0 */ +/* + * Copyright (C) Huawei Technologies Co., Ltd. 2024. All rights reserved. + */ + +#ifndef _HISI_MEM_RAS_H +#define _HISI_MEM_RAS_H + +struct hisi_mem_mbox_client_info { + struct mbox_client client; + struct pcc_mbox_chan *pcc_chan; + u64 deadline_us; + void __iomem *pcc_comm_addr; + struct completion done; +}; + +#define HISI_MEM_CAP_ACLS_EN BIT(0) +#define HISI_MEM_CAP_SPPR_EN BIT(1) +struct hisi_mem_dev { + struct platform_device *pdev; + struct hisi_mem_mbox_client_info cl_info; + struct mutex lock; + u8 chan_id; + u8 mem_type; + u8 ras_cap; +}; + +enum hisi_mem_ras_cmd_type { + HISI_MEM_RAS_CAP = 1, + HISI_MEM_RAS_ACLS = 2, + HISI_MEM_RAS_SPPR = 3, +}; + +enum hisi_mem_ras_subcmd_type { + HISI_MEM_RAS_QUERY = 0, + HISI_MEM_RAS_DO_REPAIR = 1, +}; + +enum hisi_mem_type { + MEMORY_TYPE_HBM = 0, + MEMORY_TYPE_DDR = 1, + MEMORY_TYPE_MAX = 2, +}; + +#define HISI_MEM_PCC_SHARE_MEM_BYTES_MAX 128 +#define HISI_MEM_PCC_SHARE_MEM_BYTES_MIN 64 +#define HISI_MEM_FW_INNER_HEAD_BYTES 4 +#define HISI_MEM_REQ_HEAD_BYTES 4 +#define HISI_MEM_MAX_REQ_DATA_BYTES (HISI_MEM_PCC_SHARE_MEM_BYTES_MAX - \ + HISI_MEM_FW_INNER_HEAD_BYTES - \ + HISI_MEM_REQ_HEAD_BYTES) +#define HISI_MEM_MAX_REQ_DATA_SIZE_MAX (HISI_MEM_MAX_REQ_DATA_BYTES / 4) + +#define HISI_MEM_RSP_HEAD_BYTES 4 +#define HISI_MEM_MAX_RSP_DATA_BYTES (HISI_MEM_PCC_SHARE_MEM_BYTES_MAX - \ + HISI_MEM_FW_INNER_HEAD_BYTES - \ + HISI_MEM_RSP_HEAD_BYTES) +#define HISI_MEM_MAX_RSP_DATA_SIZE_MAX (HISI_MEM_MAX_RSP_DATA_BYTES / 4) + +struct hisi_mem_req_head { + u8 subcmd; + u8 rsv[3]; +}; + +/* + * Note: Actual available size of data field also depands on the PCC header + * bytes of the specific type. Driver needs to copy the request data to the + * communication space based on the real length. + */ +struct hisi_mem_req_desc { + struct hisi_mem_req_head req_head; + u32 data[HISI_MEM_MAX_REQ_DATA_SIZE_MAX]; +}; + +#define HISI_MEM_RAS_OK 0 +#define HISI_MEM_RAS_NO_RES 1 +#define HISI_MEM_RAS_REPAIR_FAIL 2 +#define HISI_MEM_RAS_EINVAL 3 +#define HISI_MEM_RAS_ENXIO 4 +struct hisi_mem_rsp_head { + u8 ret_status; /* 0: success, other: failure */ + u8 mem_type; /* 0: HBM, 1: DDR, other: reserved */ + u8 ras_cap; /* bit0: ACLS, bit1: SPPR, other: reserved */ + u8 rsv; +}; + +/* + * Note: Actual available size of data field also depands on the PCC header + * bytes of the specific type. Driver needs to copy the response data in the + * communication space based on the real length. + */ +struct hisi_mem_rsp_desc { + struct hisi_mem_rsp_head rsp_head; + u32 data[HISI_MEM_MAX_RSP_DATA_SIZE_MAX]; +}; + +struct hisi_mem_fw_inner_head { + u8 cmd_type; + u8 msg_len; + u8 status; + u8 need_resp; +}; + +struct hisi_mem_desc { + struct hisi_mem_fw_inner_head fw_inner_head; /* 4 Bytes */ + union { + struct hisi_mem_req_desc req; + struct hisi_mem_rsp_desc rsp; + }; +}; + +#endif -- 2.33.0