diff options
Diffstat (limited to 'drivers/accel/habanalabs')
35 files changed, 1696 insertions, 1003 deletions
diff --git a/drivers/accel/habanalabs/Kconfig b/drivers/accel/habanalabs/Kconfig index be85336107f9..1919fbb169c7 100644 --- a/drivers/accel/habanalabs/Kconfig +++ b/drivers/accel/habanalabs/Kconfig @@ -6,7 +6,7 @@ config DRM_ACCEL_HABANALABS tristate "HabanaLabs AI accelerators" depends on DRM_ACCEL - depends on X86_64 + depends on X86 && X86_64 depends on PCI && HAS_IOMEM select GENERIC_ALLOCATOR select HWMON diff --git a/drivers/accel/habanalabs/common/command_submission.c b/drivers/accel/habanalabs/common/command_submission.c index 3aa6eeef443b..dee487724918 100644 --- a/drivers/accel/habanalabs/common/command_submission.c +++ b/drivers/accel/habanalabs/common/command_submission.c @@ -1360,9 +1360,8 @@ static int hl_cs_sanity_checks(struct hl_fpriv *hpriv, union hl_cs_args *args) return -EINVAL; } - if (!hl_device_operational(hdev, &status)) { + if (!hl_device_operational(hdev, &status)) return -EBUSY; - } if ((args->in.cs_flags & HL_CS_FLAGS_STAGED_SUBMISSION) && !hdev->supports_staged_submission) { @@ -2587,7 +2586,7 @@ int hl_cs_ioctl(struct drm_device *ddev, void *data, struct drm_file *file_priv) cs_seq = args->in.seq; timeout = flags & HL_CS_FLAGS_CUSTOM_TIMEOUT - ? msecs_to_jiffies(args->in.timeout * 1000) + ? secs_to_jiffies(args->in.timeout) : hpriv->hdev->timeout_jiffies; switch (cs_type) { @@ -3285,12 +3284,6 @@ static int ts_get_and_handle_kernel_record(struct hl_device *hdev, struct hl_ctx /* In case the node already registered, need to unregister first then re-use */ if (req_offset_record->ts_reg_info.in_use) { - dev_dbg(data->buf->mmg->dev, - "Requested record %p is in use on irq: %u ts addr: %p, unregister first then put on irq: %u\n", - req_offset_record, - req_offset_record->ts_reg_info.interrupt->interrupt_id, - req_offset_record->ts_reg_info.timestamp_kernel_addr, - data->interrupt->interrupt_id); /* * Since interrupt here can be different than the one the node currently registered * on, and we don't want to lock two lists while we're doing unregister, so @@ -3346,10 +3339,6 @@ static int _hl_interrupt_ts_reg_ioctl(struct hl_device *hdev, struct hl_ctx *ctx goto put_cq_cb; } - dev_dbg(hdev->dev, "Timestamp registration: interrupt id: %u, handle: 0x%llx, ts offset: %llu, cq_offset: %llu\n", - data->interrupt->interrupt_id, data->ts_handle, - data->ts_offset, data->cq_offset); - data->buf = hl_mmap_mem_buf_get(data->mmg, data->ts_handle); if (!data->buf) { rc = -EINVAL; @@ -3371,9 +3360,6 @@ static int _hl_interrupt_ts_reg_ioctl(struct hl_device *hdev, struct hl_ctx *ctx if (*pend->cq_kernel_addr >= data->target_value) { spin_unlock_irqrestore(&data->interrupt->ts_list_lock, flags); - dev_dbg(hdev->dev, "Target value already reached release ts record: pend: %p, offset: %llu, interrupt: %u\n", - pend, data->ts_offset, data->interrupt->interrupt_id); - pend->ts_reg_info.in_use = 0; *status = HL_WAIT_CS_STATUS_COMPLETED; *pend->ts_reg_info.timestamp_kernel_addr = ktime_get_ns(); diff --git a/drivers/accel/habanalabs/common/context.c b/drivers/accel/habanalabs/common/context.c index b83141f58319..9f212b17611a 100644 --- a/drivers/accel/habanalabs/common/context.c +++ b/drivers/accel/habanalabs/common/context.c @@ -199,7 +199,6 @@ out_err: int hl_ctx_init(struct hl_device *hdev, struct hl_ctx *ctx, bool is_kernel_ctx) { - char task_comm[TASK_COMM_LEN]; int rc = 0, i; ctx->hdev = hdev; @@ -272,7 +271,7 @@ int hl_ctx_init(struct hl_device *hdev, struct hl_ctx *ctx, bool is_kernel_ctx) mutex_init(&ctx->ts_reg_lock); dev_dbg(hdev->dev, "create user context, comm=\"%s\", asid=%u\n", - get_task_comm(task_comm, current), ctx->asid); + current->comm, ctx->asid); } return 0; diff --git a/drivers/accel/habanalabs/common/debugfs.c b/drivers/accel/habanalabs/common/debugfs.c index 01f071d52570..4b391807e5f2 100644 --- a/drivers/accel/habanalabs/common/debugfs.c +++ b/drivers/accel/habanalabs/common/debugfs.c @@ -42,9 +42,8 @@ static int hl_debugfs_i2c_read(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr, pkt.i2c_reg = i2c_reg; pkt.i2c_len = i2c_len; - rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), - 0, val); - if (rc) + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, val); + if (rc && rc != -EAGAIN) dev_err(hdev->dev, "Failed to read from I2C, error %d\n", rc); return rc; @@ -75,10 +74,8 @@ static int hl_debugfs_i2c_write(struct hl_device *hdev, u8 i2c_bus, u8 i2c_addr, pkt.i2c_len = i2c_len; pkt.value = cpu_to_le64(val); - rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), - 0, NULL); - - if (rc) + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); + if (rc && rc != -EAGAIN) dev_err(hdev->dev, "Failed to write to I2C, error %d\n", rc); return rc; @@ -99,10 +96,8 @@ static void hl_debugfs_led_set(struct hl_device *hdev, u8 led, u8 state) pkt.led_index = cpu_to_le32(led); pkt.value = cpu_to_le64(state); - rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), - 0, NULL); - - if (rc) + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); + if (rc && rc != -EAGAIN) dev_err(hdev->dev, "Failed to set LED %d, error %d\n", led, rc); } @@ -484,7 +479,7 @@ static ssize_t mmu_asid_va_write(struct file *file, const char __user *buf, struct hl_debugfs_entry *entry = s->private; struct hl_dbg_device_entry *dev_entry = entry->dev_entry; struct hl_device *hdev = dev_entry->hdev; - char kbuf[MMU_KBUF_SIZE]; + char kbuf[MMU_KBUF_SIZE] = {0}; char *c; ssize_t rc; @@ -546,7 +541,7 @@ static ssize_t mmu_ack_error_value_write(struct file *file, struct hl_debugfs_entry *entry = s->private; struct hl_dbg_device_entry *dev_entry = entry->dev_entry; struct hl_device *hdev = dev_entry->hdev; - char kbuf[MMU_KBUF_SIZE]; + char kbuf[MMU_KBUF_SIZE] = {0}; ssize_t rc; if (count > sizeof(kbuf) - 1) @@ -1408,7 +1403,7 @@ static ssize_t hl_timeout_locked_write(struct file *f, const char __user *buf, return rc; if (value) - hdev->timeout_jiffies = msecs_to_jiffies(value * 1000); + hdev->timeout_jiffies = secs_to_jiffies(value); else hdev->timeout_jiffies = MAX_SCHEDULE_TIMEOUT; @@ -1643,19 +1638,19 @@ static void add_files_to_device(struct hl_device *hdev, struct hl_dbg_device_ent &hl_data64b_fops); debugfs_create_file("set_power_state", - 0200, + 0644, root, dev_entry, &hl_power_fops); debugfs_create_file("device", - 0200, + 0644, root, dev_entry, &hl_device_fops); debugfs_create_file("clk_gate", - 0200, + 0644, root, dev_entry, &hl_clk_gate_fops); @@ -1667,13 +1662,13 @@ static void add_files_to_device(struct hl_device *hdev, struct hl_dbg_device_ent &hl_stop_on_err_fops); debugfs_create_file("dump_security_violations", - 0644, + 0400, root, dev_entry, &hl_security_violations_fops); debugfs_create_file("dump_razwi_events", - 0644, + 0400, root, dev_entry, &hl_razwi_check_fops); @@ -1706,7 +1701,7 @@ static void add_files_to_device(struct hl_device *hdev, struct hl_dbg_device_ent &hdev->reset_info.skip_reset_on_timeout); debugfs_create_file("state_dump", - 0600, + 0644, root, dev_entry, &hl_state_dump_fops); @@ -1722,9 +1717,14 @@ static void add_files_to_device(struct hl_device *hdev, struct hl_dbg_device_ent root, &hdev->device_release_watchdog_timeout_sec); + debugfs_create_u16("server_type", + 0444, + root, + &hdev->asic_prop.server_type); + for (i = 0, entry = dev_entry->entry_arr ; i < count ; i++, entry++) { debugfs_create_file(hl_debugfs_list[i].name, - 0444, + 0644, root, entry, &hl_debugfs_fops); diff --git a/drivers/accel/habanalabs/common/device.c b/drivers/accel/habanalabs/common/device.c index a73bd4be94b1..68eebed3b050 100644 --- a/drivers/accel/habanalabs/common/device.c +++ b/drivers/accel/habanalabs/common/device.c @@ -30,6 +30,8 @@ enum dma_alloc_type { #define MEM_SCRUB_DEFAULT_VAL 0x1122334455667788 +static void hl_device_heartbeat(struct work_struct *work); + /* * hl_set_dram_bar- sets the bar to allow later access to address * @@ -55,7 +57,8 @@ static u64 hl_set_dram_bar(struct hl_device *hdev, u64 addr, struct pci_mem_regi if (is_power_of_2(prop->dram_pci_bar_size)) bar_base_addr = addr & ~(prop->dram_pci_bar_size - 0x1ull); else - bar_base_addr = DIV_ROUND_DOWN_ULL(addr, prop->dram_pci_bar_size) * + bar_base_addr = region->region_base + + div64_u64((addr - region->region_base), prop->dram_pci_bar_size) * prop->dram_pci_bar_size; old_base = hdev->asic_funcs->set_dram_bar_base(hdev, bar_base_addr); @@ -129,8 +132,8 @@ static void *hl_dma_alloc_common(struct hl_device *hdev, size_t size, dma_addr_t } if (trace_habanalabs_dma_alloc_enabled() && !ZERO_OR_NULL_PTR(ptr)) - trace_habanalabs_dma_alloc(hdev->dev, (u64) (uintptr_t) ptr, *dma_handle, size, - caller); + trace_habanalabs_dma_alloc(&(hdev)->pdev->dev, (u64) (uintptr_t) ptr, *dma_handle, + size, caller); return ptr; } @@ -151,7 +154,7 @@ static void hl_asic_dma_free_common(struct hl_device *hdev, size_t size, void *c break; } - trace_habanalabs_dma_free(hdev->dev, store_cpu_addr, dma_handle, size, caller); + trace_habanalabs_dma_free(&(hdev)->pdev->dev, store_cpu_addr, dma_handle, size, caller); } void *hl_asic_dma_alloc_coherent_caller(struct hl_device *hdev, size_t size, dma_addr_t *dma_handle, @@ -203,15 +206,15 @@ int hl_dma_map_sgtable_caller(struct hl_device *hdev, struct sg_table *sgt, return 0; for_each_sgtable_dma_sg(sgt, sg, i) - trace_habanalabs_dma_map_page(hdev->dev, - page_to_phys(sg_page(sg)), - sg->dma_address - prop->device_dma_offset_for_host_access, + trace_habanalabs_dma_map_page(&(hdev)->pdev->dev, + page_to_phys(sg_page(sg)), + sg->dma_address - prop->device_dma_offset_for_host_access, #ifdef CONFIG_NEED_SG_DMA_LENGTH - sg->dma_length, + sg->dma_length, #else - sg->length, + sg->length, #endif - dir, caller); + dir, caller); return 0; } @@ -246,7 +249,8 @@ void hl_dma_unmap_sgtable_caller(struct hl_device *hdev, struct sg_table *sgt, if (trace_habanalabs_dma_unmap_page_enabled()) { for_each_sgtable_dma_sg(sgt, sg, i) - trace_habanalabs_dma_unmap_page(hdev->dev, page_to_phys(sg_page(sg)), + trace_habanalabs_dma_unmap_page(&(hdev)->pdev->dev, + page_to_phys(sg_page(sg)), sg->dma_address - prop->device_dma_offset_for_host_access, #ifdef CONFIG_NEED_SG_DMA_LENGTH sg->dma_length, @@ -438,16 +442,19 @@ static void print_idle_status_mask(struct hl_device *hdev, const char *message, u64 idle_mask[HL_BUSY_ENGINES_MASK_EXT_SIZE]) { if (idle_mask[3]) - dev_err(hdev->dev, "%s (mask %#llx_%016llx_%016llx_%016llx)\n", - message, idle_mask[3], idle_mask[2], idle_mask[1], idle_mask[0]); + dev_err(hdev->dev, "%s %s (mask %#llx_%016llx_%016llx_%016llx)\n", + dev_name(&hdev->pdev->dev), message, + idle_mask[3], idle_mask[2], idle_mask[1], idle_mask[0]); else if (idle_mask[2]) - dev_err(hdev->dev, "%s (mask %#llx_%016llx_%016llx)\n", - message, idle_mask[2], idle_mask[1], idle_mask[0]); + dev_err(hdev->dev, "%s %s (mask %#llx_%016llx_%016llx)\n", + dev_name(&hdev->pdev->dev), message, + idle_mask[2], idle_mask[1], idle_mask[0]); else if (idle_mask[1]) - dev_err(hdev->dev, "%s (mask %#llx_%016llx)\n", - message, idle_mask[1], idle_mask[0]); + dev_err(hdev->dev, "%s %s (mask %#llx_%016llx)\n", + dev_name(&hdev->pdev->dev), message, idle_mask[1], idle_mask[0]); else - dev_err(hdev->dev, "%s (mask %#llx)\n", message, idle_mask[0]); + dev_err(hdev->dev, "%s %s (mask %#llx)\n", dev_name(&hdev->pdev->dev), message, + idle_mask[0]); } static void hpriv_release(struct kref *ref) @@ -544,7 +551,8 @@ int hl_hpriv_put(struct hl_fpriv *hpriv) return kref_put(&hpriv->refcount, hpriv_release); } -static void print_device_in_use_info(struct hl_device *hdev, const char *message) +static void print_device_in_use_info(struct hl_device *hdev, + struct hl_mem_mgr_fini_stats *mm_fini_stats, const char *message) { u32 active_cs_num, dmabuf_export_cnt; bool unknown_reason = true; @@ -568,6 +576,12 @@ static void print_device_in_use_info(struct hl_device *hdev, const char *message dmabuf_export_cnt); } + if (mm_fini_stats->n_busy_cb) { + unknown_reason = false; + offset += scnprintf(buf + offset, size - offset, " [%u live CB handles]", + mm_fini_stats->n_busy_cb); + } + if (unknown_reason) scnprintf(buf + offset, size - offset, " [unknown reason]"); @@ -585,6 +599,7 @@ void hl_device_release(struct drm_device *ddev, struct drm_file *file_priv) { struct hl_fpriv *hpriv = file_priv->driver_priv; struct hl_device *hdev = to_hl_device(ddev); + struct hl_mem_mgr_fini_stats mm_fini_stats; if (!hdev) { pr_crit("Closing FD after device was removed. Memory leak will occur and it is advised to reboot.\n"); @@ -596,12 +611,13 @@ void hl_device_release(struct drm_device *ddev, struct drm_file *file_priv) /* Memory buffers might be still in use at this point and thus the handles IDR destruction * is postponed to hpriv_release(). */ - hl_mem_mgr_fini(&hpriv->mem_mgr); + hl_mem_mgr_fini(&hpriv->mem_mgr, &mm_fini_stats); hdev->compute_ctx_in_release = 1; if (!hl_hpriv_put(hpriv)) { - print_device_in_use_info(hdev, "User process closed FD but device still in use"); + print_device_in_use_info(hdev, &mm_fini_stats, + "User process closed FD but device still in use"); hl_device_reset(hdev, HL_DRV_RESET_HARD); } @@ -801,7 +817,7 @@ static void device_hard_reset_pending(struct work_struct *work) } queue_delayed_work(hdev->reset_wq, &device_reset_work->reset_work, - msecs_to_jiffies(HL_PENDING_RESET_PER_SEC * 1000)); + secs_to_jiffies(HL_PENDING_RESET_PER_SEC)); } } @@ -857,6 +873,10 @@ static int device_early_init(struct hl_device *hdev) gaudi2_set_asic_funcs(hdev); strscpy(hdev->asic_name, "GAUDI2C", sizeof(hdev->asic_name)); break; + case ASIC_GAUDI2D: + gaudi2_set_asic_funcs(hdev); + strscpy(hdev->asic_name, "GAUDI2D", sizeof(hdev->asic_name)); + break; default: dev_err(hdev->dev, "Unrecognized ASIC type %d\n", hdev->asic_type); @@ -945,6 +965,8 @@ static int device_early_init(struct hl_device *hdev) goto free_cb_mgr; } + INIT_DELAYED_WORK(&hdev->work_heartbeat, hl_device_heartbeat); + INIT_DELAYED_WORK(&hdev->device_reset_work.reset_work, device_hard_reset_pending); hdev->device_reset_work.hdev = hdev; hdev->device_fini_pending = 0; @@ -967,7 +989,7 @@ static int device_early_init(struct hl_device *hdev) return 0; free_cb_mgr: - hl_mem_mgr_fini(&hdev->kernel_mem_mgr); + hl_mem_mgr_fini(&hdev->kernel_mem_mgr, NULL); hl_mem_mgr_idr_destroy(&hdev->kernel_mem_mgr); free_chip_info: kfree(hdev->hl_chip_info); @@ -1011,7 +1033,7 @@ static void device_early_fini(struct hl_device *hdev) mutex_destroy(&hdev->clk_throttling.lock); - hl_mem_mgr_fini(&hdev->kernel_mem_mgr); + hl_mem_mgr_fini(&hdev->kernel_mem_mgr, NULL); hl_mem_mgr_idr_destroy(&hdev->kernel_mem_mgr); kfree(hdev->hl_chip_info); @@ -1034,31 +1056,65 @@ static void device_early_fini(struct hl_device *hdev) static bool is_pci_link_healthy(struct hl_device *hdev) { - u16 vendor_id; + u16 device_id; if (!hdev->pdev) return false; - pci_read_config_word(hdev->pdev, PCI_VENDOR_ID, &vendor_id); + pci_read_config_word(hdev->pdev, PCI_DEVICE_ID, &device_id); - return (vendor_id == PCI_VENDOR_ID_HABANALABS); + return (device_id == hdev->pdev->device); } -static int hl_device_eq_heartbeat_check(struct hl_device *hdev) +static void stringify_time_of_last_heartbeat(struct hl_device *hdev, char *time_str, size_t size, + bool is_pq_hb) { + time64_t seconds = is_pq_hb ? hdev->heartbeat_debug_info.last_pq_heartbeat_ts + : hdev->heartbeat_debug_info.last_eq_heartbeat_ts; + struct tm tm; + + if (!seconds) + return; + + time64_to_tm(seconds, 0, &tm); + + snprintf(time_str, size, "%ld-%02d-%02d %02d:%02d:%02d (UTC)", + tm.tm_year + 1900, tm.tm_mon, tm.tm_mday, tm.tm_hour, tm.tm_min, tm.tm_sec); +} + +static bool hl_device_eq_heartbeat_received(struct hl_device *hdev) +{ + struct eq_heartbeat_debug_info *heartbeat_debug_info = &hdev->heartbeat_debug_info; + u32 cpu_q_id = heartbeat_debug_info->cpu_queue_id, pq_pi_mask = (HL_QUEUE_LENGTH << 1) - 1; struct asic_fixed_properties *prop = &hdev->asic_prop; + char pq_time_str[64] = "N/A", eq_time_str[64] = "N/A"; if (!prop->cpucp_info.eq_health_check_supported) - return 0; + return true; - if (hdev->eq_heartbeat_received) { - hdev->eq_heartbeat_received = false; - } else { + if (!hdev->eq_heartbeat_received) { dev_err(hdev->dev, "EQ heartbeat event was not received!\n"); - return -EIO; + + stringify_time_of_last_heartbeat(hdev, pq_time_str, sizeof(pq_time_str), true); + stringify_time_of_last_heartbeat(hdev, eq_time_str, sizeof(eq_time_str), false); + dev_err(hdev->dev, + "EQ: {CI %u, HB counter %u, last HB time: %s}, PQ: {PI: %u, CI: %u (%u), last HB time: %s}\n", + hdev->event_queue.ci, + heartbeat_debug_info->heartbeat_event_counter, + eq_time_str, + hdev->kernel_queues[cpu_q_id].pi, + atomic_read(&hdev->kernel_queues[cpu_q_id].ci), + atomic_read(&hdev->kernel_queues[cpu_q_id].ci) & pq_pi_mask, + pq_time_str); + + hl_eq_dump(hdev, &hdev->event_queue); + + return false; } - return 0; + hdev->eq_heartbeat_received = false; + + return true; } static void hl_device_heartbeat(struct work_struct *work) @@ -1077,7 +1133,7 @@ static void hl_device_heartbeat(struct work_struct *work) * in order to validate the eq is working. * Only if both the EQ is healthy and we managed to send the next heartbeat reschedule. */ - if ((!hl_device_eq_heartbeat_check(hdev)) && (!hdev->asic_funcs->send_heartbeat(hdev))) + if (hl_device_eq_heartbeat_received(hdev) && (!hdev->asic_funcs->send_heartbeat(hdev))) goto reschedule; if (hl_device_operational(hdev, NULL)) @@ -1131,21 +1187,6 @@ static int device_late_init(struct hl_device *hdev) } hdev->high_pll = hdev->asic_prop.high_pll; - - if (hdev->heartbeat) { - /* - * Before scheduling the heartbeat driver will check if eq event has received. - * for the first schedule we need to set the indication as true then for the next - * one this indication will be true only if eq event was sent by FW. - */ - hdev->eq_heartbeat_received = true; - - INIT_DELAYED_WORK(&hdev->work_heartbeat, hl_device_heartbeat); - - schedule_delayed_work(&hdev->work_heartbeat, - usecs_to_jiffies(HL_HEARTBEAT_PER_USEC)); - } - hdev->late_init_done = true; return 0; @@ -1162,9 +1203,6 @@ static void device_late_fini(struct hl_device *hdev) if (!hdev->late_init_done) return; - if (hdev->heartbeat) - cancel_delayed_work_sync(&hdev->work_heartbeat); - if (hdev->asic_funcs->late_fini) hdev->asic_funcs->late_fini(hdev); @@ -1265,8 +1303,12 @@ static void hl_abort_waiting_for_completions(struct hl_device *hdev) static void cleanup_resources(struct hl_device *hdev, bool hard_reset, bool fw_reset, bool skip_wq_flush) { - if (hard_reset) + if (hard_reset) { + if (hdev->heartbeat) + cancel_delayed_work_sync(&hdev->work_heartbeat); + device_late_fini(hdev); + } /* * Halt the engines and disable interrupts so we won't get any more @@ -1494,15 +1536,14 @@ static void send_disable_pci_access(struct hl_device *hdev, u32 flags) * of heartbeat, the device CPU is marked as disable * so this message won't be sent */ - if (hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0)) { - dev_warn(hdev->dev, "Failed to disable FW's PCI access\n"); + if (hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0)) return; - } - /* verify that last EQs are handled before disabled is set */ + /* disable_irq also generates sync irq, this verifies that last EQs are handled + * before disabled is set. The IRQ will be enabled again in request_irq call. + */ if (hdev->cpu_queues_enable) - synchronize_irq(pci_irq_vector(hdev->pdev, - hdev->asic_prop.eq_interrupt_id)); + disable_irq(pci_irq_vector(hdev->pdev, hdev->asic_prop.eq_interrupt_id)); } } @@ -1546,6 +1587,31 @@ static void handle_reset_trigger(struct hl_device *hdev, u32 flags) } } +static void reset_heartbeat_debug_info(struct hl_device *hdev) +{ + hdev->heartbeat_debug_info.last_pq_heartbeat_ts = 0; + hdev->heartbeat_debug_info.last_eq_heartbeat_ts = 0; + hdev->heartbeat_debug_info.heartbeat_event_counter = 0; +} + +static inline void device_heartbeat_schedule(struct hl_device *hdev) +{ + if (!hdev->heartbeat) + return; + + reset_heartbeat_debug_info(hdev); + + /* + * Before scheduling the heartbeat driver will check if eq event has received. + * for the first schedule we need to set the indication as true then for the next + * one this indication will be true only if eq event was sent by FW. + */ + hdev->eq_heartbeat_received = true; + + schedule_delayed_work(&hdev->work_heartbeat, + usecs_to_jiffies(HL_HEARTBEAT_PER_USEC)); +} + /* * hl_device_reset - reset the device * @@ -1768,14 +1834,16 @@ kill_processes: hdev->device_cpu_disabled = false; hdev->reset_info.hard_reset_pending = false; + /* + * Put the device in an unusable state if there are 2 back to back resets due to + * fatal errors. + */ if (hdev->reset_info.reset_trigger_repeated && - (hdev->reset_info.prev_reset_trigger == - HL_DRV_RESET_FW_FATAL_ERR)) { - /* if there 2 back to back resets from FW, - * ensure driver puts the driver in a unusable state - */ + (hdev->reset_info.prev_reset_trigger == HL_DRV_RESET_FW_FATAL_ERR || + hdev->reset_info.prev_reset_trigger == + HL_DRV_RESET_HEARTBEAT)) { dev_crit(hdev->dev, - "%s Consecutive FW fatal errors received, stopping hard reset\n", + "%s Consecutive fatal errors, stopping hard reset\n", dev_name(&(hdev)->pdev->dev)); rc = -EIO; goto out_err; @@ -1913,6 +1981,8 @@ kill_processes: if (hard_reset) { hdev->reset_info.hard_reset_cnt++; + device_heartbeat_schedule(hdev); + /* After reset is done, we are ready to receive events from * the F/W. We can't do it before because we will ignore events * and if those events are fatal, we won't know about it and @@ -2021,7 +2091,7 @@ int hl_device_cond_reset(struct hl_device *hdev, u32 flags, u64 event_mask) dev_dbg(hdev->dev, "Device is going to be hard-reset in %u sec unless being released\n", hdev->device_release_watchdog_timeout_sec); schedule_delayed_work(&hdev->device_release_watchdog_work.reset_work, - msecs_to_jiffies(hdev->device_release_watchdog_timeout_sec * 1000)); + secs_to_jiffies(hdev->device_release_watchdog_timeout_sec)); hdev->reset_info.watchdog_active = 1; out: spin_unlock(&hdev->reset_info.lock); @@ -2347,6 +2417,12 @@ int hl_device_init(struct hl_device *hdev) goto out_disabled; } + /* Scheduling the EQ heartbeat thread must come after driver is done with all + * initializations, as we want to make sure the FW gets enough time to be prepared + * to respond to heartbeat packets. + */ + device_heartbeat_schedule(hdev); + dev_notice(hdev->dev, "Successfully added device %s to habanalabs driver\n", dev_name(&(hdev)->pdev->dev)); @@ -2589,7 +2665,7 @@ inline u32 hl_rreg(struct hl_device *hdev, u32 reg) u32 val = readl(hdev->rmmio + reg); if (unlikely(trace_habanalabs_rreg32_enabled())) - trace_habanalabs_rreg32(hdev->dev, reg, val); + trace_habanalabs_rreg32(&(hdev)->pdev->dev, reg, val); return val; } @@ -2607,7 +2683,7 @@ inline u32 hl_rreg(struct hl_device *hdev, u32 reg) inline void hl_wreg(struct hl_device *hdev, u32 reg, u32 val) { if (unlikely(trace_habanalabs_wreg32_enabled())) - trace_habanalabs_wreg32(hdev->dev, reg, val); + trace_habanalabs_wreg32(&(hdev)->pdev->dev, reg, val); writel(val, hdev->rmmio + reg); } @@ -2801,3 +2877,88 @@ void hl_enable_err_info_capture(struct hl_error_info *captured_err_info) atomic_set(&captured_err_info->cs_timeout.write_enable, 1); captured_err_info->undef_opcode.write_enable = true; } + +void hl_init_cpu_for_irq(struct hl_device *hdev) +{ +#ifdef CONFIG_NUMA + struct cpumask *available_mask = &hdev->irq_affinity_mask; + int numa_node = hdev->pdev->dev.numa_node, i; + static struct cpumask cpu_mask; + + if (numa_node < 0) + return; + + if (!cpumask_and(&cpu_mask, cpumask_of_node(numa_node), cpu_online_mask)) { + dev_err(hdev->dev, "No available affinities in current numa node\n"); + return; + } + + /* Remove HT siblings */ + for_each_cpu(i, &cpu_mask) + cpumask_set_cpu(cpumask_first(topology_sibling_cpumask(i)), available_mask); +#endif +} + +void hl_set_irq_affinity(struct hl_device *hdev, int irq) +{ + if (cpumask_empty(&hdev->irq_affinity_mask)) { + dev_dbg(hdev->dev, "affinity mask is empty\n"); + return; + } + + if (irq_set_affinity_and_hint(irq, &hdev->irq_affinity_mask)) + dev_err(hdev->dev, "Failed setting irq %d affinity\n", irq); +} + +void hl_eq_heartbeat_event_handle(struct hl_device *hdev) +{ + hdev->heartbeat_debug_info.heartbeat_event_counter++; + hdev->heartbeat_debug_info.last_eq_heartbeat_ts = ktime_get_real_seconds(); + hdev->eq_heartbeat_received = true; +} + +void hl_handle_clk_change_event(struct hl_device *hdev, u16 event_type, u64 *event_mask) +{ + struct hl_clk_throttle *clk_throttle = &hdev->clk_throttling; + ktime_t zero_time = ktime_set(0, 0); + + mutex_lock(&clk_throttle->lock); + + switch (event_type) { + case EQ_EVENT_POWER_EVT_START: + clk_throttle->current_reason |= HL_CLK_THROTTLE_POWER; + clk_throttle->aggregated_reason |= HL_CLK_THROTTLE_POWER; + clk_throttle->timestamp[HL_CLK_THROTTLE_TYPE_POWER].start = ktime_get(); + clk_throttle->timestamp[HL_CLK_THROTTLE_TYPE_POWER].end = zero_time; + dev_dbg_ratelimited(hdev->dev, "Clock throttling due to power consumption\n"); + break; + + case EQ_EVENT_POWER_EVT_END: + clk_throttle->current_reason &= ~HL_CLK_THROTTLE_POWER; + clk_throttle->timestamp[HL_CLK_THROTTLE_TYPE_POWER].end = ktime_get(); + dev_dbg_ratelimited(hdev->dev, "Power envelop is safe, back to optimal clock\n"); + break; + + case EQ_EVENT_THERMAL_EVT_START: + clk_throttle->current_reason |= HL_CLK_THROTTLE_THERMAL; + clk_throttle->aggregated_reason |= HL_CLK_THROTTLE_THERMAL; + clk_throttle->timestamp[HL_CLK_THROTTLE_TYPE_THERMAL].start = ktime_get(); + clk_throttle->timestamp[HL_CLK_THROTTLE_TYPE_THERMAL].end = zero_time; + *event_mask |= HL_NOTIFIER_EVENT_USER_ENGINE_ERR; + dev_info_ratelimited(hdev->dev, "Clock throttling due to overheating\n"); + break; + + case EQ_EVENT_THERMAL_EVT_END: + clk_throttle->current_reason &= ~HL_CLK_THROTTLE_THERMAL; + clk_throttle->timestamp[HL_CLK_THROTTLE_TYPE_THERMAL].end = ktime_get(); + *event_mask |= HL_NOTIFIER_EVENT_USER_ENGINE_ERR; + dev_info_ratelimited(hdev->dev, "Thermal envelop is safe, back to optimal clock\n"); + break; + + default: + dev_err(hdev->dev, "Received invalid clock change event %d\n", event_type); + break; + } + + mutex_unlock(&clk_throttle->lock); +} diff --git a/drivers/accel/habanalabs/common/firmware_if.c b/drivers/accel/habanalabs/common/firmware_if.c index 3558a6a8e192..eeb6b2a80fc7 100644 --- a/drivers/accel/habanalabs/common/firmware_if.c +++ b/drivers/accel/habanalabs/common/firmware_if.c @@ -8,6 +8,7 @@ #include "habanalabs.h" #include <linux/habanalabs/hl_boot_if.h> +#include <linux/pci.h> #include <linux/firmware.h> #include <linux/crc32.h> #include <linux/slab.h> @@ -40,6 +41,31 @@ static char *comms_sts_str_arr[COMMS_STS_INVLD_LAST] = { [COMMS_STS_TIMEOUT_ERR] = __stringify(COMMS_STS_TIMEOUT_ERR), }; +/** + * hl_fw_version_cmp() - compares the FW version to a specific version + * + * @hdev: pointer to hl_device structure + * @major: major number of a reference version + * @minor: minor number of a reference version + * @subminor: sub-minor number of a reference version + * + * Return 1 if FW version greater than the reference version, -1 if it's + * smaller and 0 if versions are identical. + */ +int hl_fw_version_cmp(struct hl_device *hdev, u32 major, u32 minor, u32 subminor) +{ + if (hdev->fw_sw_major_ver != major) + return (hdev->fw_sw_major_ver > major) ? 1 : -1; + + if (hdev->fw_sw_minor_ver != minor) + return (hdev->fw_sw_minor_ver > minor) ? 1 : -1; + + if (hdev->fw_sw_sub_minor_ver != subminor) + return (hdev->fw_sw_sub_minor_ver > subminor) ? 1 : -1; + + return 0; +} + static char *extract_fw_ver_from_str(const char *fw_str) { char *str, *fw_ver, *whitespace; @@ -345,43 +371,63 @@ int hl_fw_load_fw_to_device(struct hl_device *hdev, const char *fw_name, int hl_fw_send_pci_access_msg(struct hl_device *hdev, u32 opcode, u64 value) { struct cpucp_packet pkt = {}; + int rc; pkt.ctl = cpu_to_le32(opcode << CPUCP_PKT_CTL_OPCODE_SHIFT); pkt.value = cpu_to_le64(value); - return hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); + if (rc) + dev_err(hdev->dev, "Failed to disable FW's PCI access\n"); + + return rc; } +/** + * hl_fw_send_cpu_message() - send CPU message to the device. + * + * @hdev: pointer to hl_device structure. + * @hw_queue_id: HW queue ID + * @msg: raw data of the message/packet + * @size: size of @msg in bytes + * @timeout_us: timeout in usec to wait for CPU reply on the message + * @result: return code reported by FW + * + * send message to the device CPU. + * + * Return: 0 on success, non-zero for failure. + * -ENOMEM: memory allocation failure + * -EAGAIN: CPU is disabled (try again when enabled) + * -ETIMEDOUT: timeout waiting for FW response + * -EIO: protocol error + */ int hl_fw_send_cpu_message(struct hl_device *hdev, u32 hw_queue_id, u32 *msg, - u16 len, u32 timeout, u64 *result) + u16 size, u32 timeout_us, u64 *result) { struct hl_hw_queue *queue = &hdev->kernel_queues[hw_queue_id]; struct asic_fixed_properties *prop = &hdev->asic_prop; + u32 tmp, expected_ack_val, pi, opcode; struct cpucp_packet *pkt; dma_addr_t pkt_dma_addr; struct hl_bd *sent_bd; - u32 tmp, expected_ack_val, pi, opcode; - int rc; + int rc = 0, fw_rc; - pkt = hl_cpu_accessible_dma_pool_alloc(hdev, len, &pkt_dma_addr); + pkt = hl_cpu_accessible_dma_pool_alloc(hdev, size, &pkt_dma_addr); if (!pkt) { - dev_err(hdev->dev, - "Failed to allocate DMA memory for packet to CPU\n"); + dev_err(hdev->dev, "Failed to allocate DMA memory for packet to CPU\n"); return -ENOMEM; } - memcpy(pkt, msg, len); + memcpy(pkt, msg, size); mutex_lock(&hdev->send_cpu_message_lock); /* CPU-CP messages can be sent during soft-reset */ - if (hdev->disabled && !hdev->reset_info.in_compute_reset) { - rc = 0; + if (hdev->disabled && !hdev->reset_info.in_compute_reset) goto out; - } if (hdev->device_cpu_disabled) { - rc = -EIO; + rc = -EAGAIN; goto out; } @@ -397,7 +443,7 @@ int hl_fw_send_cpu_message(struct hl_device *hdev, u32 hw_queue_id, u32 *msg, * Which means that we don't need to lock the access to the entire H/W * queues module when submitting a JOB to the CPU queue. */ - hl_hw_queue_submit_bd(hdev, queue, hl_queue_inc_ptr(queue->pi), len, pkt_dma_addr); + hl_hw_queue_submit_bd(hdev, queue, hl_queue_inc_ptr(queue->pi), size, pkt_dma_addr); if (prop->fw_app_cpu_boot_dev_sts0 & CPU_BOOT_DEV_STS0_PKT_PI_ACK_EN) expected_ack_val = queue->pi; @@ -406,7 +452,7 @@ int hl_fw_send_cpu_message(struct hl_device *hdev, u32 hw_queue_id, u32 *msg, rc = hl_poll_timeout_memory(hdev, &pkt->fence, tmp, (tmp == expected_ack_val), 1000, - timeout, true); + timeout_us, true); hl_hw_queue_inc_ci_kernel(hdev, hw_queue_id); @@ -414,19 +460,27 @@ int hl_fw_send_cpu_message(struct hl_device *hdev, u32 hw_queue_id, u32 *msg, /* If FW performed reset just before sending it a packet, we will get a timeout. * This is expected behavior, hence no need for error message. */ - if (!hl_device_operational(hdev, NULL) && !hdev->reset_info.in_compute_reset) + if (!hl_device_operational(hdev, NULL) && !hdev->reset_info.in_compute_reset) { dev_dbg(hdev->dev, "Device CPU packet timeout (0x%x) due to FW reset\n", tmp); - else - dev_err(hdev->dev, "Device CPU packet timeout (status = 0x%x)\n", tmp); + } else { + struct hl_bd *bd = queue->kernel_address; + + bd += hl_pi_2_offset(pi); + + dev_err(hdev->dev, "Device CPU packet timeout (status = 0x%x)\n" + "Pkt info[%u]: dma_addr: 0x%llx, kernel_addr: %p, len:0x%x, ctl: 0x%x, ptr:0x%llx, dram_bd:%u\n", + tmp, pi, pkt_dma_addr, (void *)pkt, bd->len, bd->ctl, bd->ptr, + queue->dram_bd); + } hdev->device_cpu_disabled = true; goto out; } tmp = le32_to_cpu(pkt->ctl); - rc = (tmp & CPUCP_PKT_CTL_RC_MASK) >> CPUCP_PKT_CTL_RC_SHIFT; - if (rc) { + fw_rc = (tmp & CPUCP_PKT_CTL_RC_MASK) >> CPUCP_PKT_CTL_RC_SHIFT; + if (fw_rc) { opcode = (tmp & CPUCP_PKT_CTL_OPCODE_MASK) >> CPUCP_PKT_CTL_OPCODE_SHIFT; if (!prop->supports_advanced_cpucp_rc) { @@ -435,7 +489,7 @@ int hl_fw_send_cpu_message(struct hl_device *hdev, u32 hw_queue_id, u32 *msg, goto scrub_descriptor; } - switch (rc) { + switch (fw_rc) { case cpucp_packet_invalid: dev_err(hdev->dev, "CPU packet %d is not supported by F/W\n", opcode); @@ -460,7 +514,7 @@ int hl_fw_send_cpu_message(struct hl_device *hdev, u32 hw_queue_id, u32 *msg, /* propagate the return code from the f/w to the callers who want to check it */ if (result) - *result = rc; + *result = fw_rc; rc = -EIO; @@ -480,7 +534,7 @@ scrub_descriptor: out: mutex_unlock(&hdev->send_cpu_message_lock); - hl_cpu_accessible_dma_pool_free(hdev, len, pkt); + hl_cpu_accessible_dma_pool_free(hdev, size, pkt); return rc; } @@ -501,7 +555,7 @@ int hl_fw_unmask_irq(struct hl_device *hdev, u16 event_type) 0, &result); if (rc) - dev_err(hdev->dev, "failed to unmask RAZWI IRQ %d", event_type); + dev_err(hdev->dev, "failed to unmask event %d", event_type); return rc; } @@ -540,7 +594,7 @@ int hl_fw_unmask_irq_arr(struct hl_device *hdev, const u32 *irq_arr, total_pkt_size, 0, &result); if (rc) - dev_err(hdev->dev, "failed to unmask IRQ array\n"); + dev_err(hdev->dev, "failed to unmask event array\n"); kfree(pkt); @@ -550,7 +604,7 @@ int hl_fw_unmask_irq_arr(struct hl_device *hdev, const u32 *irq_arr, int hl_fw_test_cpu_queue(struct hl_device *hdev) { struct cpucp_packet test_pkt = {}; - u64 result; + u64 result = 0; int rc; test_pkt.ctl = cpu_to_le32(CPUCP_PACKET_TEST << @@ -623,16 +677,14 @@ int hl_fw_send_device_activity(struct hl_device *hdev, bool open) int hl_fw_send_heartbeat(struct hl_device *hdev) { struct cpucp_packet hb_pkt; - u64 result; + u64 result = 0; int rc; memset(&hb_pkt, 0, sizeof(hb_pkt)); - hb_pkt.ctl = cpu_to_le32(CPUCP_PACKET_TEST << - CPUCP_PKT_CTL_OPCODE_SHIFT); + hb_pkt.ctl = cpu_to_le32(CPUCP_PACKET_TEST << CPUCP_PKT_CTL_OPCODE_SHIFT); hb_pkt.value = cpu_to_le64(CPUCP_PACKET_FENCE_VAL); - rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &hb_pkt, - sizeof(hb_pkt), 0, &result); + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &hb_pkt, sizeof(hb_pkt), 0, &result); if ((rc) || (result != CPUCP_PACKET_FENCE_VAL)) return -EIO; @@ -643,6 +695,8 @@ int hl_fw_send_heartbeat(struct hl_device *hdev) rc = -EIO; } + hdev->heartbeat_debug_info.last_pq_heartbeat_ts = ktime_get_real_seconds(); + return rc; } @@ -885,7 +939,7 @@ static int hl_fw_send_msi_info_msg(struct hl_device *hdev) { struct cpucp_array_data_packet *pkt; size_t total_pkt_size, data_size; - u64 result; + u64 result = 0; int rc; /* skip sending this info for unsupported ASICs */ @@ -976,11 +1030,10 @@ int hl_fw_get_eeprom_data(struct hl_device *hdev, void *data, size_t max_size) rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_EEPROM_TIMEOUT_USEC, &result); - if (rc) { - dev_err(hdev->dev, - "Failed to handle CPU-CP EEPROM packet, error %d\n", - rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP EEPROM packet, error %d\n", rc); goto out; } @@ -1021,7 +1074,9 @@ int hl_fw_get_monitor_dump(struct hl_device *hdev, void *data) rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_MON_DUMP_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, "Failed to handle CPU-CP monitor-dump packet, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP monitor-dump packet, error %d\n", rc); goto out; } @@ -1055,8 +1110,9 @@ int hl_fw_cpucp_pci_counters_get(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, - "Failed to handle CPU-CP PCI info pkt, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP PCI info pkt, error %d\n", rc); return rc; } counters->rx_throughput = result; @@ -1070,8 +1126,9 @@ int hl_fw_cpucp_pci_counters_get(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, - "Failed to handle CPU-CP PCI info pkt, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP PCI info pkt, error %d\n", rc); return rc; } counters->tx_throughput = result; @@ -1084,8 +1141,9 @@ int hl_fw_cpucp_pci_counters_get(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, - "Failed to handle CPU-CP PCI info pkt, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP PCI info pkt, error %d\n", rc); return rc; } counters->replay_cnt = (u32) result; @@ -1105,9 +1163,9 @@ int hl_fw_cpucp_total_energy_get(struct hl_device *hdev, u64 *total_energy) rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, - "Failed to handle CpuCP total energy pkt, error %d\n", - rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CpuCP total energy pkt, error %d\n", rc); return rc; } @@ -1183,7 +1241,8 @@ int hl_fw_cpucp_pll_info_get(struct hl_device *hdev, u32 pll_index, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, "Failed to read PLL info, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, "Failed to read PLL info, error %d\n", rc); return rc; } @@ -1210,7 +1269,8 @@ int hl_fw_cpucp_power_get(struct hl_device *hdev, u64 *power) rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, "Failed to read power, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, "Failed to read power, error %d\n", rc); return rc; } @@ -1247,8 +1307,9 @@ int hl_fw_dram_replaced_row_get(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); if (rc) { - dev_err(hdev->dev, - "Failed to handle CPU-CP replaced rows info pkt, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP replaced rows info pkt, error %d\n", rc); goto out; } @@ -1273,7 +1334,8 @@ int hl_fw_dram_pending_row_get(struct hl_device *hdev, u32 *pend_rows_num) rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, &result); if (rc) { - dev_err(hdev->dev, + if (rc != -EAGAIN) + dev_err(hdev->dev, "Failed to handle CPU-CP pending rows info pkt, error %d\n", rc); goto out; } @@ -1428,7 +1490,7 @@ int hl_fw_wait_preboot_ready(struct hl_device *hdev) { struct pre_fw_load_props *pre_fw_load = &hdev->fw_loader.pre_fw_load; u32 status = 0, timeout; - int rc, tries = 1; + int rc, tries = 1, fw_err = 0; bool preboot_still_runs; /* Need to check two possible scenarios: @@ -1468,18 +1530,18 @@ retry: } } - if (rc) { + /* If we read all FF, then something is totally wrong, no point + * of reading specific errors + */ + if (status != -1) + fw_err = fw_read_errors(hdev, pre_fw_load->boot_err0_reg, + pre_fw_load->boot_err1_reg, + pre_fw_load->sts_boot_dev_sts0_reg, + pre_fw_load->sts_boot_dev_sts1_reg); + if (rc || fw_err) { detect_cpu_boot_status(hdev, status); - dev_err(hdev->dev, "CPU boot ready timeout (status = %d)\n", status); - - /* If we read all FF, then something is totally wrong, no point - * of reading specific errors - */ - if (status != -1) - fw_read_errors(hdev, pre_fw_load->boot_err0_reg, - pre_fw_load->boot_err1_reg, - pre_fw_load->sts_boot_dev_sts0_reg, - pre_fw_load->sts_boot_dev_sts1_reg); + dev_err(hdev->dev, "CPU boot %s (status = %d)\n", + fw_err ? "failed due to an error" : "ready timeout", status); return -EIO; } @@ -1750,7 +1812,7 @@ static void hl_fw_dynamic_send_cmd(struct hl_device *hdev, val = FIELD_PREP(COMMS_COMMAND_CMD_MASK, cmd); val |= FIELD_PREP(COMMS_COMMAND_SIZE_MASK, size); - trace_habanalabs_comms_send_cmd(hdev->dev, comms_cmd_str_arr[cmd]); + trace_habanalabs_comms_send_cmd(&hdev->pdev->dev, comms_cmd_str_arr[cmd]); WREG32(le32_to_cpu(dyn_regs->kmd_msg_to_cpu), val); } @@ -1808,7 +1870,7 @@ static int hl_fw_dynamic_wait_for_status(struct hl_device *hdev, dyn_regs = &fw_loader->dynamic_loader.comm_desc.cpu_dyn_regs; - trace_habanalabs_comms_wait_status(hdev->dev, comms_sts_str_arr[expected_status]); + trace_habanalabs_comms_wait_status(&hdev->pdev->dev, comms_sts_str_arr[expected_status]); /* Wait for expected status */ rc = hl_poll_timeout( @@ -1825,7 +1887,8 @@ static int hl_fw_dynamic_wait_for_status(struct hl_device *hdev, return -EIO; } - trace_habanalabs_comms_wait_status_done(hdev->dev, comms_sts_str_arr[expected_status]); + trace_habanalabs_comms_wait_status_done(&hdev->pdev->dev, + comms_sts_str_arr[expected_status]); /* * skip storing FW response for NOOP to preserve the actual desired @@ -1899,7 +1962,7 @@ int hl_fw_dynamic_send_protocol_cmd(struct hl_device *hdev, { int rc; - trace_habanalabs_comms_protocol_cmd(hdev->dev, comms_cmd_str_arr[cmd]); + trace_habanalabs_comms_protocol_cmd(&hdev->pdev->dev, comms_cmd_str_arr[cmd]); /* first send clear command to clean former commands */ rc = hl_fw_dynamic_send_clear_cmd(hdev, fw_loader); @@ -2038,7 +2101,7 @@ static int hl_fw_dynamic_validate_descriptor(struct hl_device *hdev, * note that no alignment/stride address issues here as all structures * are 64 bit padded. */ - data_ptr = (u8 *)fw_desc + sizeof(struct comms_desc_header); + data_ptr = (u8 *)fw_desc + sizeof(struct comms_msg_header); data_size = le16_to_cpu(fw_desc->header.size); data_crc32 = hl_fw_compat_crc32(data_ptr, data_size); @@ -2192,11 +2255,11 @@ static int hl_fw_dynamic_read_and_validate_descriptor(struct hl_device *hdev, memcpy_fromio(fw_desc, src, sizeof(struct lkd_fw_comms_desc)); fw_data_size = le16_to_cpu(fw_desc->header.size); - temp_fw_desc = vzalloc(sizeof(struct comms_desc_header) + fw_data_size); + temp_fw_desc = vzalloc(sizeof(struct comms_msg_header) + fw_data_size); if (!temp_fw_desc) return -ENOMEM; - memcpy_fromio(temp_fw_desc, src, sizeof(struct comms_desc_header) + fw_data_size); + memcpy_fromio(temp_fw_desc, src, sizeof(struct comms_msg_header) + fw_data_size); rc = hl_fw_dynamic_validate_descriptor(hdev, fw_loader, (struct lkd_fw_comms_desc *) temp_fw_desc); @@ -2718,18 +2781,20 @@ static int hl_fw_dynamic_init_cpu(struct hl_device *hdev, hdev->reset_info.curr_reset_cause = HL_RESET_CAUSE_UNKNOWN; } + rc = hl_fw_dynamic_request_descriptor(hdev, fw_loader, sizeof(struct lkd_msg_comms)); + if (rc) + goto protocol_err; + + if (hdev->asic_prop.support_dynamic_resereved_fw_size) + hdev->asic_prop.reserved_fw_mem_size = + le32_to_cpu(fw_loader->dynamic_loader.comm_desc.rsvd_mem_size_mb) * SZ_1M; + if (!(hdev->fw_components & FW_TYPE_BOOT_CPU)) { struct lkd_fw_binning_info *binning_info; - rc = hl_fw_dynamic_request_descriptor(hdev, fw_loader, - sizeof(struct lkd_msg_comms)); - if (rc) - goto protocol_err; - /* read preboot version */ rc = hl_fw_dynamic_read_device_fw_version(hdev, FW_COMP_PREBOOT, fw_loader->dynamic_loader.comm_desc.cur_fw_ver); - if (rc) return rc; @@ -2756,11 +2821,6 @@ static int hl_fw_dynamic_init_cpu(struct hl_device *hdev, hdev->decoder_binning, hdev->rotator_binning); } - if (hdev->asic_prop.support_dynamic_resereved_fw_size) { - hdev->asic_prop.reserved_fw_mem_size = - le32_to_cpu(fw_loader->dynamic_loader.comm_desc.rsvd_mem_size_mb); - } - return 0; } @@ -2795,7 +2855,7 @@ static int hl_fw_dynamic_init_cpu(struct hl_device *hdev, hdev->asic_funcs->init_cpu_scrambler_dram(hdev); if (!(hdev->fw_components & FW_TYPE_LINUX)) { - dev_info(hdev->dev, "Skip loading Linux F/W\n"); + dev_dbg(hdev->dev, "Skip loading Linux F/W\n"); return 0; } @@ -3125,10 +3185,10 @@ long hl_fw_get_frequency(struct hl_device *hdev, u32 pll_index, bool curr) pkt.pll_index = cpu_to_le32((u32)used_pll_idx); rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, &result); - if (rc) { - dev_err(hdev->dev, "Failed to get frequency of PLL %d, error %d\n", - used_pll_idx, rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, "Failed to get frequency of PLL %d, error %d\n", + used_pll_idx, rc); return rc; } @@ -3152,8 +3212,7 @@ void hl_fw_set_frequency(struct hl_device *hdev, u32 pll_index, u64 freq) pkt.value = cpu_to_le64(freq); rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); - - if (rc) + if (rc && rc != -EAGAIN) dev_err(hdev->dev, "Failed to set frequency to PLL %d, error %d\n", used_pll_idx, rc); } @@ -3169,9 +3228,9 @@ long hl_fw_get_max_power(struct hl_device *hdev) pkt.ctl = cpu_to_le32(CPUCP_PACKET_MAX_POWER_GET << CPUCP_PKT_CTL_OPCODE_SHIFT); rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, &result); - if (rc) { - dev_err(hdev->dev, "Failed to get max power, error %d\n", rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, "Failed to get max power, error %d\n", rc); return rc; } @@ -3193,8 +3252,7 @@ void hl_fw_set_max_power(struct hl_device *hdev) pkt.value = cpu_to_le64(hdev->max_power); rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); - - if (rc) + if (rc && rc != -EAGAIN) dev_err(hdev->dev, "Failed to set max power, error %d\n", rc); } @@ -3220,11 +3278,11 @@ static int hl_fw_get_sec_attest_data(struct hl_device *hdev, u32 packet_id, void pkt.data_max_size = cpu_to_le32(size); pkt.nonce = cpu_to_le32(nonce); - rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), - timeout, NULL); + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), timeout, NULL); if (rc) { - dev_err(hdev->dev, - "Failed to handle CPU-CP pkt %u, error %d\n", packet_id, rc); + if (rc != -EAGAIN) + dev_err(hdev->dev, + "Failed to handle CPU-CP pkt %u, error %d\n", packet_id, rc); goto out; } @@ -3266,10 +3324,12 @@ int hl_fw_send_generic_request(struct hl_device *hdev, enum hl_passthrough_type rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *)&pkt, sizeof(pkt), HL_CPUCP_INFO_TIMEOUT_USEC, &result); - if (rc) - dev_err(hdev->dev, "failed to send CPUCP data of generic fw pkt\n"); - else + if (rc) { + if (rc != -EAGAIN) + dev_err(hdev->dev, "failed to send CPUCP data of generic fw pkt\n"); + } else { dev_dbg(hdev->dev, "generic pkt was successful, result: 0x%llx\n", result); + } *size = (u32)result; diff --git a/drivers/accel/habanalabs/common/habanalabs.h b/drivers/accel/habanalabs/common/habanalabs.h index 2a900c9941fe..6f27ce4fa01b 100644 --- a/drivers/accel/habanalabs/common/habanalabs.h +++ b/drivers/accel/habanalabs/common/habanalabs.h @@ -71,7 +71,7 @@ struct hl_fpriv; #define HL_DEVICE_TIMEOUT_USEC 1000000 /* 1 s */ -#define HL_HEARTBEAT_PER_USEC 5000000 /* 5 s */ +#define HL_HEARTBEAT_PER_USEC 10000000 /* 10 s */ #define HL_PLL_LOW_JOB_FREQ_USEC 5000000 /* 5 s */ @@ -443,18 +443,22 @@ enum hl_collective_mode { * a CB handle can be provided for jobs on this queue. * Otherwise, a CB address must be provided. * @collective_mode: collective mode of current queue + * @q_dram_bd_address: PQ dram address, used when PQ need to reside in DRAM. * @driver_only: true if only the driver is allowed to send a job to this queue, * false otherwise. * @binned: True if the queue is binned out and should not be used * @supports_sync_stream: True if queue supports sync stream + * @dram_bd: True if the bd should be copied to dram, needed for PQ which has been allocated on dram */ struct hw_queue_properties { enum hl_queue_type type; enum queue_cb_alloc_flags cb_alloc_flags; enum hl_collective_mode collective_mode; + u64 q_dram_bd_address; u8 driver_only; u8 binned; u8 supports_sync_stream; + u8 dram_bd; }; /** @@ -590,8 +594,6 @@ struct hl_hints_range { * we display to the user * @mmu_pgt_size: MMU page tables total size. * @mmu_pte_size: PTE size in MMU page tables. - * @mmu_hop_table_size: MMU hop table size. - * @mmu_hop0_tables_total_size: total size of MMU hop0 tables. * @dram_page_size: The DRAM physical page size. * @cfg_size: configuration space size on SRAM. * @sram_size: total size of SRAM. @@ -645,10 +647,12 @@ struct hl_hints_range { * @num_engine_cores: number of engine cpu cores. * @max_num_of_engines: maximum number of all engines in the ASIC. * @num_of_special_blocks: special_blocks array size. - * @glbl_err_cause_num: global err cause number. + * @glbl_err_max_cause_num: global err max cause number. * @hbw_flush_reg: register to read to generate HBW flush. value of 0 means HBW flush is * not supported. - * @reserved_fw_mem_size: size in MB of dram memory reserved for FW. + * @reserved_fw_mem_size: size of dram memory reserved for FW. + * @fw_event_queue_size: queue size for events from CPU-CP. + * A value of 0 means using the default HL_EQ_SIZE_IN_BYTES value. * @collective_first_sob: first sync object available for collective use * @collective_first_mon: first monitor available for collective use * @sync_stream_first_sob: first sync object available for sync stream use @@ -743,8 +747,6 @@ struct asic_fixed_properties { u32 clk_pll_index; u32 mmu_pgt_size; u32 mmu_pte_size; - u32 mmu_hop_table_size; - u32 mmu_hop0_tables_total_size; u32 dram_page_size; u32 cfg_size; u32 sram_size; @@ -779,9 +781,10 @@ struct asic_fixed_properties { u32 num_engine_cores; u32 max_num_of_engines; u32 num_of_special_blocks; - u32 glbl_err_cause_num; + u32 glbl_err_max_cause_num; u32 hbw_flush_reg; u32 reserved_fw_mem_size; + u32 fw_event_queue_size; u16 collective_first_sob; u16 collective_first_mon; u16 sync_stream_first_sob; @@ -902,6 +905,18 @@ struct hl_mem_mgr { }; /** + * struct hl_mem_mgr_fini_stats - describes statistics returned during memory manager teardown. + * @n_busy_cb: the amount of CB handles that could not be removed + * @n_busy_ts: the amount of TS handles that could not be removed + * @n_busy_other: the amount of any other type of handles that could not be removed + */ +struct hl_mem_mgr_fini_stats { + u32 n_busy_cb; + u32 n_busy_ts; + u32 n_busy_other; +}; + +/** * struct hl_mmap_mem_buf_behavior - describes unified memory manager buffer behavior * @topic: string identifier used for logging * @mem_id: memory type identifier, embedded in the handle and used to identify @@ -1052,6 +1067,8 @@ struct hl_encaps_signals_mgr { * @collective_mode: collective mode of current queue * @kernel_address: holds the queue's kernel virtual address. * @bus_address: holds the queue's DMA address. + * @pq_dram_address: hold the dram address when the PQ is allocated, used when dram_bd is true in + * queue properites. * @pi: holds the queue's pi value. * @ci: holds the queue's ci value, AS CALCULATED BY THE DRIVER (not real ci). * @hw_queue_id: the id of the H/W queue. @@ -1061,6 +1078,7 @@ struct hl_encaps_signals_mgr { * @valid: is the queue valid (we have array of 32 queues, not all of them * exist). * @supports_sync_stream: True if queue supports sync stream + * @dram_bd: True if the bd should be copied to dram, needed for PQ which has been allocated on dram */ struct hl_hw_queue { struct hl_cs_job **shadow_queue; @@ -1069,6 +1087,7 @@ struct hl_hw_queue { enum hl_collective_mode collective_mode; void *kernel_address; dma_addr_t bus_address; + u64 pq_dram_address; u32 pi; atomic_t ci; u32 hw_queue_id; @@ -1077,6 +1096,7 @@ struct hl_hw_queue { u16 int_queue_len; u8 valid; u8 supports_sync_stream; + u8 dram_bd; }; /** @@ -1224,6 +1244,7 @@ struct hl_user_pending_interrupt { * @hdev: pointer to the device structure * @kernel_address: holds the queue's kernel virtual address * @bus_address: holds the queue's DMA address + * @size: the event queue size * @ci: ci inside the queue * @prev_eqe_index: the index of the previous event queue entry. The index of * the current entry's index must be +1 of the previous one. @@ -1235,6 +1256,7 @@ struct hl_eq { struct hl_device *hdev; void *kernel_address; dma_addr_t bus_address; + u32 size; u32 ci; u32 prev_eqe_index; bool check_eqe_index; @@ -1263,15 +1285,18 @@ struct hl_dec { * @ASIC_GAUDI2: Gaudi2 device. * @ASIC_GAUDI2B: Gaudi2B device. * @ASIC_GAUDI2C: Gaudi2C device. + * @ASIC_GAUDI2D: Gaudi2D device. */ enum hl_asic_type { ASIC_INVALID, + ASIC_GOYA, ASIC_GAUDI, ASIC_GAUDI_SEC, ASIC_GAUDI2, ASIC_GAUDI2B, ASIC_GAUDI2C, + ASIC_GAUDI2D, }; struct hl_cs_parser; @@ -2547,7 +2572,7 @@ struct hl_state_dump_specs { * DEVICES */ -#define HL_STR_MAX 32 +#define HL_STR_MAX 64 #define HL_DEV_STS_MAX (HL_DEVICE_STATUS_LAST + 1) @@ -2704,11 +2729,16 @@ void hl_wreg(struct hl_device *hdev, u32 reg, u32 val); * updated directly by the device. If false, the host memory being polled will * be updated by host CPU. Required so host knows whether or not the memory * might need to be byte-swapped before returning value to caller. + * + * On the first 4 polling iterations the macro goes to sleep for short period of + * time that gradually increases and reaches sleep_us on the fifth iteration. */ #define hl_poll_timeout_memory(hdev, addr, val, cond, sleep_us, timeout_us, \ mem_written_by_device) \ ({ \ + u64 __sleep_step_us; \ ktime_t __timeout; \ + u8 __step = 8; \ \ __timeout = ktime_add_us(ktime_get(), timeout_us); \ might_sleep_if(sleep_us); \ @@ -2726,8 +2756,10 @@ void hl_wreg(struct hl_device *hdev, u32 reg, u32 val); (val) = le32_to_cpu(*(__le32 *) &(val)); \ break; \ } \ - if (sleep_us) \ - usleep_range((sleep_us >> 2) + 1, sleep_us); \ + __sleep_step_us = sleep_us >> __step; \ + if (__sleep_step_us) \ + usleep_range((__sleep_step_us >> 2) + 1, __sleep_step_us); \ + __step >>= 1; \ } \ (cond) ? 0 : -ETIMEDOUT; \ }) @@ -3170,6 +3202,21 @@ struct hl_reset_info { }; /** + * struct eq_heartbeat_debug_info - stores debug info to be used upon heartbeat failure. + * @last_pq_heartbeat_ts: timestamp of the last test packet that was sent to FW. + * This packet is the trigger in FW to send the EQ heartbeat event. + * @last_eq_heartbeat_ts: timestamp of the last EQ heartbeat event that was received from FW. + * @heartbeat_event_counter: number of heartbeat events received. + * @cpu_queue_id: used to read the queue pi/ci + */ +struct eq_heartbeat_debug_info { + time64_t last_pq_heartbeat_ts; + time64_t last_eq_heartbeat_ts; + u32 heartbeat_event_counter; + u32 cpu_queue_id; +}; + +/** * struct hl_device - habanalabs device structure. * @pdev: pointer to PCI device, can be NULL in case of simulator device. * @pcie_bar_phys: array of available PCIe bars physical addresses. @@ -3257,6 +3304,8 @@ struct hl_reset_info { * @clk_throttling: holds information about current/previous clock throttling events * @captured_err_info: holds information about errors. * @reset_info: holds current device reset information. + * @heartbeat_debug_info: counters used to debug heartbeat failures. + * @irq_affinity_mask: mask of available CPU cores for user and decoder interrupt handling. * @stream_master_qid_arr: pointer to array with QIDs of master streams. * @fw_inner_major_ver: the major of current loaded preboot inner version. * @fw_inner_minor_ver: the minor of current loaded preboot inner version. @@ -3446,6 +3495,10 @@ struct hl_device { struct hl_reset_info reset_info; + struct eq_heartbeat_debug_info heartbeat_debug_info; + + cpumask_t irq_affinity_mask; + u32 *stream_master_qid_arr; u32 fw_inner_major_ver; u32 fw_inner_minor_ver; @@ -3588,25 +3641,6 @@ struct hl_ioctl_desc { hl_ioctl_t *func; }; -static inline bool hl_is_fw_sw_ver_below(struct hl_device *hdev, u32 fw_sw_major, u32 fw_sw_minor) -{ - if (hdev->fw_sw_major_ver < fw_sw_major) - return true; - if (hdev->fw_sw_major_ver > fw_sw_major) - return false; - if (hdev->fw_sw_minor_ver < fw_sw_minor) - return true; - return false; -} - -static inline bool hl_is_fw_sw_ver_equal_or_greater(struct hl_device *hdev, u32 fw_sw_major, - u32 fw_sw_minor) -{ - return (hdev->fw_sw_major_ver > fw_sw_major || - (hdev->fw_sw_major_ver == fw_sw_major && - hdev->fw_sw_minor_ver >= fw_sw_minor)); -} - /* * Kernel module functions that can be accessed by entire module */ @@ -3732,6 +3766,7 @@ int hl_eq_init(struct hl_device *hdev, struct hl_eq *q); void hl_eq_fini(struct hl_device *hdev, struct hl_eq *q); void hl_cq_reset(struct hl_device *hdev, struct hl_cq *q); void hl_eq_reset(struct hl_device *hdev, struct hl_eq *q); +void hl_eq_dump(struct hl_device *hdev, struct hl_eq *q); irqreturn_t hl_irq_handler_cq(int irq, void *arg); irqreturn_t hl_irq_handler_eq(int irq, void *arg); irqreturn_t hl_irq_handler_dec_abnrm(int irq, void *arg); @@ -3886,6 +3921,7 @@ int hl_mmu_hr_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, struct hl_mmu_hop_ struct hl_hr_mmu_funcs *hr_func); int hl_mmu_if_set_funcs(struct hl_device *hdev); void hl_mmu_v1_set_funcs(struct hl_device *hdev, struct hl_mmu_funcs *mmu); +void hl_mmu_v2_set_funcs(struct hl_device *hdev, struct hl_mmu_funcs *mmu); void hl_mmu_v2_hr_set_funcs(struct hl_device *hdev, struct hl_mmu_funcs *mmu); int hl_mmu_va_to_pa(struct hl_ctx *ctx, u64 virt_addr, u64 *phys_addr); int hl_mmu_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, @@ -3893,7 +3929,24 @@ int hl_mmu_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, u64 hl_mmu_scramble_addr(struct hl_device *hdev, u64 addr); u64 hl_mmu_descramble_addr(struct hl_device *hdev, u64 addr); bool hl_is_dram_va(struct hl_device *hdev, u64 virt_addr); - +struct pgt_info *hl_mmu_dr_get_pgt_info(struct hl_ctx *ctx, u64 hop_addr); +void hl_mmu_dr_free_hop(struct hl_ctx *ctx, u64 hop_addr); +void hl_mmu_dr_free_pgt_node(struct hl_ctx *ctx, struct pgt_info *pgt_info); +u64 hl_mmu_dr_get_phys_hop0_addr(struct hl_ctx *ctx); +u64 hl_mmu_dr_get_hop0_addr(struct hl_ctx *ctx); +void hl_mmu_dr_write_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, u64 val); +void hl_mmu_dr_write_final_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, u64 val); +void hl_mmu_dr_clear_pte(struct hl_ctx *ctx, u64 pte_addr); +u64 hl_mmu_dr_get_phys_addr(struct hl_ctx *ctx, u64 shadow_addr); +void hl_mmu_dr_get_pte(struct hl_ctx *ctx, u64 hop_addr); +int hl_mmu_dr_put_pte(struct hl_ctx *ctx, u64 hop_addr); +u64 hl_mmu_dr_get_alloc_next_hop_addr(struct hl_ctx *ctx, u64 curr_pte, bool *is_new_hop); +u64 hl_mmu_dr_alloc_hop(struct hl_ctx *ctx); +void hl_mmu_dr_flush(struct hl_ctx *ctx); +int hl_mmu_dr_init(struct hl_device *hdev); +void hl_mmu_dr_fini(struct hl_device *hdev); + +int hl_fw_version_cmp(struct hl_device *hdev, u32 major, u32 minor, u32 subminor); int hl_fw_load_fw_to_device(struct hl_device *hdev, const char *fw_name, void __iomem *dst, u32 src_offset, u32 size); int hl_fw_send_pci_access_msg(struct hl_device *hdev, u32 opcode, u64 value); @@ -4008,7 +4061,7 @@ char *hl_format_as_binary(char *buf, size_t buf_len, u32 n); const char *hl_sync_engine_to_string(enum hl_sync_engine_type engine_type); void hl_mem_mgr_init(struct device *dev, struct hl_mem_mgr *mmg); -void hl_mem_mgr_fini(struct hl_mem_mgr *mmg); +void hl_mem_mgr_fini(struct hl_mem_mgr *mmg, struct hl_mem_mgr_fini_stats *stats); void hl_mem_mgr_idr_destroy(struct hl_mem_mgr *mmg); int hl_mem_mgr_mmap(struct hl_mem_mgr *mmg, struct vm_area_struct *vma, void *args); @@ -4032,6 +4085,10 @@ void hl_handle_critical_hw_err(struct hl_device *hdev, u16 event_id, u64 *event_ void hl_handle_fw_err(struct hl_device *hdev, struct hl_info_fw_err_info *info); void hl_capture_engine_err(struct hl_device *hdev, u16 engine_id, u16 error_count); void hl_enable_err_info_capture(struct hl_error_info *captured_err_info); +void hl_init_cpu_for_irq(struct hl_device *hdev); +void hl_set_irq_affinity(struct hl_device *hdev, int irq); +void hl_eq_heartbeat_event_handle(struct hl_device *hdev); +void hl_handle_clk_change_event(struct hl_device *hdev, u16 event_type, u64 *event_mask); #ifdef CONFIG_DEBUG_FS diff --git a/drivers/accel/habanalabs/common/habanalabs_drv.c b/drivers/accel/habanalabs/common/habanalabs_drv.c index e542fd40e16c..0035748f3228 100644 --- a/drivers/accel/habanalabs/common/habanalabs_drv.c +++ b/drivers/accel/habanalabs/common/habanalabs_drv.c @@ -101,7 +101,6 @@ static const struct drm_driver hl_driver = { .major = LINUX_VERSION_MAJOR, .minor = LINUX_VERSION_PATCHLEVEL, .patchlevel = LINUX_VERSION_SUBLEVEL, - .date = "20190505", .fops = &hl_fops, .open = hl_device_open, @@ -144,6 +143,9 @@ static enum hl_asic_type get_asic_type(struct hl_device *hdev) case REV_ID_C: asic_type = ASIC_GAUDI2C; break; + case REV_ID_D: + asic_type = ASIC_GAUDI2D; + break; default: break; } @@ -260,7 +262,7 @@ int hl_device_open(struct drm_device *ddev, struct drm_file *file_priv) out_err: mutex_unlock(&hdev->fpriv_list_lock); - hl_mem_mgr_fini(&hpriv->mem_mgr); + hl_mem_mgr_fini(&hpriv->mem_mgr, NULL); hl_mem_mgr_idr_destroy(&hpriv->mem_mgr); hl_ctx_mgr_fini(hpriv->hdev, &hpriv->ctx_mgr); mutex_destroy(&hpriv->ctx_lock); @@ -359,8 +361,7 @@ static void fixup_device_params_per_asic(struct hl_device *hdev, int timeout) * a different default timeout for Gaudi */ if (timeout == HL_DEFAULT_TIMEOUT_LOCKED) - hdev->timeout_jiffies = msecs_to_jiffies(GAUDI_DEFAULT_TIMEOUT_LOCKED * - MSEC_PER_SEC); + hdev->timeout_jiffies = secs_to_jiffies(GAUDI_DEFAULT_TIMEOUT_LOCKED); hdev->reset_upon_device_release = 0; break; @@ -385,7 +386,7 @@ static int fixup_device_params(struct hl_device *hdev) hdev->fw_comms_poll_interval_usec = HL_FW_STATUS_POLL_INTERVAL_USEC; if (tmp_timeout) - hdev->timeout_jiffies = msecs_to_jiffies(tmp_timeout * MSEC_PER_SEC); + hdev->timeout_jiffies = secs_to_jiffies(tmp_timeout); else hdev->timeout_jiffies = MAX_SCHEDULE_TIMEOUT; diff --git a/drivers/accel/habanalabs/common/habanalabs_ioctl.c b/drivers/accel/habanalabs/common/habanalabs_ioctl.c index 1dd6e23172ca..dc80ca921d90 100644 --- a/drivers/accel/habanalabs/common/habanalabs_ioctl.c +++ b/drivers/accel/habanalabs/common/habanalabs_ioctl.c @@ -17,8 +17,6 @@ #include <linux/uaccess.h> #include <linux/vmalloc.h> -#include <asm/msr.h> - /* make sure there is space for all the signed info */ static_assert(sizeof(struct cpucp_info) <= SEC_DEV_INFO_BUF_SZ); @@ -1279,13 +1277,10 @@ static long _hl_ioctl(struct hl_fpriv *hpriv, unsigned int cmd, unsigned long ar retcode = -EFAULT; out_err: - if (retcode) { - char task_comm[TASK_COMM_LEN]; - + if (retcode) dev_dbg_ratelimited(dev, "error in ioctl: pid=%d, comm=\"%s\", cmd=%#010x, nr=%#04x\n", - task_pid_nr(current), get_task_comm(task_comm, current), cmd, nr); - } + task_pid_nr(current), current->comm, cmd, nr); if (kdata != stack_kdata) kfree(kdata); @@ -1308,11 +1303,9 @@ long hl_ioctl_control(struct file *filep, unsigned int cmd, unsigned long arg) if (nr == _IOC_NR(DRM_IOCTL_HL_INFO)) { ioctl = &hl_ioctls_control[nr - HL_COMMAND_START]; } else { - char task_comm[TASK_COMM_LEN]; - dev_dbg_ratelimited(hdev->dev_ctrl, "invalid ioctl: pid=%d, comm=\"%s\", cmd=%#010x, nr=%#04x\n", - task_pid_nr(current), get_task_comm(task_comm, current), cmd, nr); + task_pid_nr(current), current->comm, cmd, nr); return -ENOTTY; } diff --git a/drivers/accel/habanalabs/common/hw_queue.c b/drivers/accel/habanalabs/common/hw_queue.c index d0087c0ec48c..3d04a7507cce 100644 --- a/drivers/accel/habanalabs/common/hw_queue.c +++ b/drivers/accel/habanalabs/common/hw_queue.c @@ -84,6 +84,8 @@ void hl_hw_queue_submit_bd(struct hl_device *hdev, struct hl_hw_queue *q, u32 ctl, u32 len, u64 ptr) { struct hl_bd *bd; + u64 addr; + int i; bd = q->kernel_address; bd += hl_pi_2_offset(q->pi); @@ -91,7 +93,16 @@ void hl_hw_queue_submit_bd(struct hl_device *hdev, struct hl_hw_queue *q, bd->len = cpu_to_le32(len); bd->ptr = cpu_to_le64(ptr); + if (q->dram_bd) + for (i = 0 ; i < 2 ; i++) { + addr = q->pq_dram_address + + ((hl_pi_2_offset(q->pi) * sizeof(struct hl_bd)) + (i * sizeof(u64))); + hdev->asic_funcs->access_dev_mem(hdev, PCI_REGION_DRAM, addr, + (u64 *)(bd) + i, DEBUGFS_WRITE64); + } + q->pi = hl_queue_inc_ptr(q->pi); + hdev->asic_funcs->ring_doorbell(hdev, q->hw_queue_id, q->pi); } @@ -1087,12 +1098,18 @@ int hl_hw_queues_create(struct hl_device *hdev) q->supports_sync_stream = asic->hw_queues_props[i].supports_sync_stream; q->collective_mode = asic->hw_queues_props[i].collective_mode; + q->dram_bd = asic->hw_queues_props[i].dram_bd; + rc = queue_init(hdev, q, i); if (rc) { dev_err(hdev->dev, "failed to initialize queue %d\n", i); goto release_queues; } + + /* Set DRAM PQ address for the queue if it should be at DRAM */ + if (q->dram_bd) + q->pq_dram_address = asic->hw_queues_props[i].q_dram_bd_address; } return 0; diff --git a/drivers/accel/habanalabs/common/hwmon.c b/drivers/accel/habanalabs/common/hwmon.c index 1ee2ee07e9ed..52d1e6bf10dc 100644 --- a/drivers/accel/habanalabs/common/hwmon.c +++ b/drivers/accel/habanalabs/common/hwmon.c @@ -46,7 +46,7 @@ static u32 fixup_flags_legacy_fw(struct hl_device *hdev, enum hwmon_sensor_types break; default: - dev_err(hdev->dev, "unsupported h/w sensor type %d\n", type); + dev_err_ratelimited(hdev->dev, "unsupported h/w sensor type %d\n", type); flags = cpucp_flags; break; } @@ -134,7 +134,7 @@ static u32 adjust_hwmon_flags(struct hl_device *hdev, enum hwmon_sensor_types ty break; default: - dev_err(hdev->dev, "unsupported h/w sensor type %d\n", type); + dev_err_ratelimited(hdev->dev, "unsupported h/w sensor type %d\n", type); flags = cpucp_flags; break; } @@ -162,7 +162,8 @@ int hl_build_hwmon_channel_info(struct hl_device *hdev, struct cpucp_sensor *sen break; if (type >= HWMON_NR_SENSOR_TYPES) { - dev_err(hdev->dev, "Got wrong sensor type %d from device\n", type); + dev_err_ratelimited(hdev->dev, + "Got wrong sensor type %d from device\n", type); return -EINVAL; } @@ -584,9 +585,10 @@ int hl_get_temperature(struct hl_device *hdev, *value = (long) result; if (rc) { - dev_err(hdev->dev, - "Failed to get temperature from sensor %d, error %d\n", - sensor_index, rc); + if (rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, + "Failed to get temperature from sensor %d, error %d\n", + sensor_index, rc); *value = 0; } @@ -609,9 +611,8 @@ int hl_set_temperature(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); - - if (rc) - dev_err(hdev->dev, + if (rc && rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, "Failed to set temperature of sensor %d, error %d\n", sensor_index, rc); @@ -638,9 +639,10 @@ int hl_get_voltage(struct hl_device *hdev, *value = (long) result; if (rc) { - dev_err(hdev->dev, - "Failed to get voltage from sensor %d, error %d\n", - sensor_index, rc); + if (rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, + "Failed to get voltage from sensor %d, error %d\n", + sensor_index, rc); *value = 0; } @@ -667,9 +669,10 @@ int hl_get_current(struct hl_device *hdev, *value = (long) result; if (rc) { - dev_err(hdev->dev, - "Failed to get current from sensor %d, error %d\n", - sensor_index, rc); + if (rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, + "Failed to get current from sensor %d, error %d\n", + sensor_index, rc); *value = 0; } @@ -696,9 +699,10 @@ int hl_get_fan_speed(struct hl_device *hdev, *value = (long) result; if (rc) { - dev_err(hdev->dev, - "Failed to get fan speed from sensor %d, error %d\n", - sensor_index, rc); + if (rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, + "Failed to get fan speed from sensor %d, error %d\n", + sensor_index, rc); *value = 0; } @@ -725,9 +729,10 @@ int hl_get_pwm_info(struct hl_device *hdev, *value = (long) result; if (rc) { - dev_err(hdev->dev, - "Failed to get pwm info from sensor %d, error %d\n", - sensor_index, rc); + if (rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, + "Failed to get pwm info from sensor %d, error %d\n", + sensor_index, rc); *value = 0; } @@ -750,9 +755,8 @@ void hl_set_pwm_info(struct hl_device *hdev, int sensor_index, u32 attr, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); - - if (rc) - dev_err(hdev->dev, + if (rc && rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, "Failed to set pwm info to sensor %d, error %d\n", sensor_index, rc); } @@ -773,9 +777,8 @@ int hl_set_voltage(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); - - if (rc) - dev_err(hdev->dev, + if (rc && rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, "Failed to set voltage of sensor %d, error %d\n", sensor_index, rc); @@ -796,11 +799,9 @@ int hl_set_current(struct hl_device *hdev, pkt.type = __cpu_to_le16(attr); pkt.value = __cpu_to_le64(value); - rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), - 0, NULL); - - if (rc) - dev_err(hdev->dev, + rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); + if (rc && rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, "Failed to set current of sensor %d, error %d\n", sensor_index, rc); @@ -829,9 +830,8 @@ int hl_set_power(struct hl_device *hdev, rc = hdev->asic_funcs->send_cpu_message(hdev, (u32 *) &pkt, sizeof(pkt), 0, NULL); - - if (rc) - dev_err(hdev->dev, + if (rc && rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, "Failed to set power of sensor %d, error %d\n", sensor_index, rc); @@ -858,9 +858,10 @@ int hl_get_power(struct hl_device *hdev, *value = (long) result; if (rc) { - dev_err(hdev->dev, - "Failed to get power of sensor %d, error %d\n", - sensor_index, rc); + if (rc != -EAGAIN) + dev_err_ratelimited(hdev->dev, + "Failed to get power of sensor %d, error %d\n", + sensor_index, rc); *value = 0; } diff --git a/drivers/accel/habanalabs/common/irq.c b/drivers/accel/habanalabs/common/irq.c index 978b7f4d5eeb..7c9f2f6a2870 100644 --- a/drivers/accel/habanalabs/common/irq.c +++ b/drivers/accel/habanalabs/common/irq.c @@ -652,14 +652,16 @@ void hl_cq_reset(struct hl_device *hdev, struct hl_cq *q) */ int hl_eq_init(struct hl_device *hdev, struct hl_eq *q) { + u32 size = hdev->asic_prop.fw_event_queue_size ? : HL_EQ_SIZE_IN_BYTES; void *p; - p = hl_cpu_accessible_dma_pool_alloc(hdev, HL_EQ_SIZE_IN_BYTES, &q->bus_address); + p = hl_cpu_accessible_dma_pool_alloc(hdev, size, &q->bus_address); if (!p) return -ENOMEM; q->hdev = hdev; q->kernel_address = p; + q->size = size; q->ci = 0; q->prev_eqe_index = 0; @@ -678,7 +680,7 @@ void hl_eq_fini(struct hl_device *hdev, struct hl_eq *q) { flush_workqueue(hdev->eq_wq); - hl_cpu_accessible_dma_pool_free(hdev, HL_EQ_SIZE_IN_BYTES, q->kernel_address); + hl_cpu_accessible_dma_pool_free(hdev, q->size, q->kernel_address); } void hl_eq_reset(struct hl_device *hdev, struct hl_eq *q) @@ -693,5 +695,30 @@ void hl_eq_reset(struct hl_device *hdev, struct hl_eq *q) * when the device is operational again */ - memset(q->kernel_address, 0, HL_EQ_SIZE_IN_BYTES); + memset(q->kernel_address, 0, q->size); +} + +void hl_eq_dump(struct hl_device *hdev, struct hl_eq *q) +{ + u32 eq_length, eqe_size, ctl, ready, mode, type, index; + struct hl_eq_header *hdr; + u8 *ptr; + int i; + + eq_length = HL_EQ_LENGTH; + eqe_size = q->size / HL_EQ_LENGTH; + + dev_info(hdev->dev, "Contents of EQ entries headers:\n"); + + for (i = 0, ptr = q->kernel_address ; i < eq_length ; ++i, ptr += eqe_size) { + hdr = (struct hl_eq_header *) ptr; + ctl = le32_to_cpu(hdr->ctl); + ready = FIELD_GET(EQ_CTL_READY_MASK, ctl); + mode = FIELD_GET(EQ_CTL_EVENT_MODE_MASK, ctl); + type = FIELD_GET(EQ_CTL_EVENT_TYPE_MASK, ctl); + index = FIELD_GET(EQ_CTL_INDEX_MASK, ctl); + + dev_info(hdev->dev, "%02u: %#010x [ready: %u, mode %u, type %04u, index %05u]\n", + i, ctl, ready, mode, type, index); + } } diff --git a/drivers/accel/habanalabs/common/memory.c b/drivers/accel/habanalabs/common/memory.c index 3348ad12c237..601fdbe70179 100644 --- a/drivers/accel/habanalabs/common/memory.c +++ b/drivers/accel/habanalabs/common/memory.c @@ -14,7 +14,7 @@ #include <linux/vmalloc.h> #include <linux/pci-p2pdma.h> -MODULE_IMPORT_NS(DMA_BUF); +MODULE_IMPORT_NS("DMA_BUF"); #define HL_MMU_DEBUG 0 diff --git a/drivers/accel/habanalabs/common/memory_mgr.c b/drivers/accel/habanalabs/common/memory_mgr.c index c4d84df355b0..99cd83139d46 100644 --- a/drivers/accel/habanalabs/common/memory_mgr.c +++ b/drivers/accel/habanalabs/common/memory_mgr.c @@ -318,28 +318,61 @@ void hl_mem_mgr_init(struct device *dev, struct hl_mem_mgr *mmg) idr_init(&mmg->handles); } +static void hl_mem_mgr_fini_stats_reset(struct hl_mem_mgr_fini_stats *stats) +{ + if (!stats) + return; + + memset(stats, 0, sizeof(*stats)); +} + +static void hl_mem_mgr_fini_stats_inc(u64 mem_id, struct hl_mem_mgr_fini_stats *stats) +{ + if (!stats) + return; + + switch (mem_id) { + case HL_MMAP_TYPE_CB: + ++stats->n_busy_cb; + break; + case HL_MMAP_TYPE_TS_BUFF: + ++stats->n_busy_ts; + break; + default: + /* we currently store only CB/TS so this shouldn't happen */ + ++stats->n_busy_other; + } +} + /** * hl_mem_mgr_fini - release unified memory manager * * @mmg: parent unified memory manager + * @stats: if non-NULL, will return some counters for handles that could not be removed. * * Release the unified memory manager. Shall be called from an interrupt context. */ -void hl_mem_mgr_fini(struct hl_mem_mgr *mmg) +void hl_mem_mgr_fini(struct hl_mem_mgr *mmg, struct hl_mem_mgr_fini_stats *stats) { struct hl_mmap_mem_buf *buf; struct idr *idp; const char *topic; + u64 mem_id; u32 id; + hl_mem_mgr_fini_stats_reset(stats); + idp = &mmg->handles; idr_for_each_entry(idp, buf, id) { topic = buf->behavior->topic; - if (hl_mmap_mem_buf_put(buf) != 1) + mem_id = buf->behavior->mem_id; + if (hl_mmap_mem_buf_put(buf) != 1) { dev_err(mmg->dev, "%s: Buff handle %u for CTX is still alive\n", topic, id); + hl_mem_mgr_fini_stats_inc(mem_id, stats); + } } } diff --git a/drivers/accel/habanalabs/common/mmu/Makefile b/drivers/accel/habanalabs/common/mmu/Makefile index 1806c524e04a..f4b815bf4f7d 100644 --- a/drivers/accel/habanalabs/common/mmu/Makefile +++ b/drivers/accel/habanalabs/common/mmu/Makefile @@ -1,3 +1,3 @@ # SPDX-License-Identifier: GPL-2.0-only HL_COMMON_MMU_FILES := common/mmu/mmu.o common/mmu/mmu_v1.o \ - common/mmu/mmu_v2_hr.o + common/mmu/mmu_v2.o common/mmu/mmu_v2_hr.o diff --git a/drivers/accel/habanalabs/common/mmu/mmu.c b/drivers/accel/habanalabs/common/mmu/mmu.c index b654302a68fc..79823facce7f 100644 --- a/drivers/accel/habanalabs/common/mmu/mmu.c +++ b/drivers/accel/habanalabs/common/mmu/mmu.c @@ -6,6 +6,7 @@ */ #include <linux/slab.h> +#include <linux/pci.h> #include "../habanalabs.h" @@ -262,7 +263,7 @@ int hl_mmu_unmap_page(struct hl_ctx *ctx, u64 virt_addr, u32 page_size, bool flu mmu_funcs->flush(ctx); if (trace_habanalabs_mmu_unmap_enabled() && !rc) - trace_habanalabs_mmu_unmap(hdev->dev, virt_addr, 0, page_size, flush_pte); + trace_habanalabs_mmu_unmap(&hdev->pdev->dev, virt_addr, 0, page_size, flush_pte); return rc; } @@ -349,7 +350,7 @@ int hl_mmu_map_page(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, u32 page_s if (flush_pte) mmu_funcs->flush(ctx); - trace_habanalabs_mmu_map(hdev->dev, virt_addr, phys_addr, page_size, flush_pte); + trace_habanalabs_mmu_map(&hdev->pdev->dev, virt_addr, phys_addr, page_size, flush_pte); return 0; @@ -585,6 +586,8 @@ int hl_mmu_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, int hl_mmu_if_set_funcs(struct hl_device *hdev) { + struct asic_fixed_properties *prop = &hdev->asic_prop; + if (hdev->mmu_disable) return 0; @@ -597,8 +600,10 @@ int hl_mmu_if_set_funcs(struct hl_device *hdev) case ASIC_GAUDI2: case ASIC_GAUDI2B: case ASIC_GAUDI2C: - /* MMUs in Gaudi2 are always host resident */ - hl_mmu_v2_hr_set_funcs(hdev, &hdev->mmu_func[MMU_HR_PGT]); + case ASIC_GAUDI2D: + hl_mmu_v2_set_funcs(hdev, &hdev->mmu_func[MMU_DR_PGT]); + if (prop->pmmu.host_resident) + hl_mmu_v2_hr_set_funcs(hdev, &hdev->mmu_func[MMU_HR_PGT]); break; default: dev_err(hdev->dev, "Unrecognized ASIC type %d\n", @@ -641,7 +646,8 @@ int hl_mmu_invalidate_cache(struct hl_device *hdev, bool is_hard, u32 flags) rc = hdev->asic_funcs->mmu_invalidate_cache(hdev, is_hard, flags); if (rc) dev_err_ratelimited(hdev->dev, - "%s cache invalidation failed, rc=%d\n", + "%s: %s cache invalidation failed, rc=%d\n", + dev_name(&hdev->pdev->dev), flags == VM_TYPE_USERPTR ? "PMMU" : "HMMU", rc); return rc; @@ -656,8 +662,9 @@ int hl_mmu_invalidate_cache_range(struct hl_device *hdev, bool is_hard, asid, va, size); if (rc) dev_err_ratelimited(hdev->dev, - "%s cache range invalidation failed: va=%#llx, size=%llu, rc=%d", - flags == VM_TYPE_USERPTR ? "PMMU" : "HMMU", va, size, rc); + "%s: %s cache range invalidation failed: va=%#llx, size=%llu, rc=%d", + dev_name(&hdev->pdev->dev), flags == VM_TYPE_USERPTR ? "PMMU" : "HMMU", + va, size, rc); return rc; } @@ -1209,3 +1216,219 @@ int hl_mmu_hr_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, struct hl_mmu_hop_ return 0; } +struct pgt_info *hl_mmu_dr_get_pgt_info(struct hl_ctx *ctx, u64 hop_addr) +{ + struct pgt_info *pgt_info = NULL; + + hash_for_each_possible(ctx->mmu_shadow_hash, pgt_info, node, + (unsigned long) hop_addr) + if (hop_addr == pgt_info->shadow_addr) + break; + + return pgt_info; +} + +void hl_mmu_dr_free_hop(struct hl_ctx *ctx, u64 hop_addr) +{ + struct pgt_info *pgt_info = hl_mmu_dr_get_pgt_info(ctx, hop_addr); + + hl_mmu_dr_free_pgt_node(ctx, pgt_info); +} + +void hl_mmu_dr_free_pgt_node(struct hl_ctx *ctx, struct pgt_info *pgt_info) +{ + struct hl_device *hdev = ctx->hdev; + + gen_pool_free(hdev->mmu_priv.dr.mmu_pgt_pool, pgt_info->phys_addr, + hdev->asic_prop.dmmu.hop_table_size); + hash_del(&pgt_info->node); + kfree((u64 *) (uintptr_t) pgt_info->shadow_addr); + kfree(pgt_info); +} + +u64 hl_mmu_dr_get_phys_hop0_addr(struct hl_ctx *ctx) +{ + return ctx->hdev->asic_prop.mmu_pgt_addr + + (ctx->asid * ctx->hdev->asic_prop.dmmu.hop_table_size); +} + +u64 hl_mmu_dr_get_hop0_addr(struct hl_ctx *ctx) +{ + return (u64) (uintptr_t) ctx->hdev->mmu_priv.dr.mmu_shadow_hop0 + + (ctx->asid * ctx->hdev->asic_prop.dmmu.hop_table_size); +} + +u64 hl_mmu_dr_get_phys_addr(struct hl_ctx *ctx, u64 shadow_addr) +{ + u64 page_mask = ctx->hdev->asic_prop.dmmu.hop_table_size - 1; + u64 shadow_hop_addr = shadow_addr & (~page_mask); + u64 pte_offset = shadow_addr & page_mask; + u64 phys_hop_addr; + + if (shadow_hop_addr != hl_mmu_dr_get_hop0_addr(ctx)) + phys_hop_addr = hl_mmu_dr_get_pgt_info(ctx, shadow_hop_addr)->phys_addr; + else + phys_hop_addr = hl_mmu_dr_get_phys_hop0_addr(ctx); + + return phys_hop_addr + pte_offset; +} + +void hl_mmu_dr_write_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, u64 val) +{ + u64 phys_val = hl_mmu_dr_get_phys_addr(ctx, val); + + ctx->hdev->asic_funcs->write_pte(ctx->hdev, hl_mmu_dr_get_phys_addr(ctx, shadow_pte_addr), + phys_val); + + *(u64 *) (uintptr_t) shadow_pte_addr = val; +} + +void hl_mmu_dr_write_final_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, u64 val) +{ + ctx->hdev->asic_funcs->write_pte(ctx->hdev, + hl_mmu_dr_get_phys_addr(ctx, shadow_pte_addr), val); + *(u64 *) (uintptr_t) shadow_pte_addr = val; +} + +void hl_mmu_dr_clear_pte(struct hl_ctx *ctx, u64 pte_addr) +{ + hl_mmu_dr_write_final_pte(ctx, pte_addr, 0); +} + +void hl_mmu_dr_get_pte(struct hl_ctx *ctx, u64 hop_addr) +{ + hl_mmu_dr_get_pgt_info(ctx, hop_addr)->num_of_ptes++; +} + +int hl_mmu_dr_put_pte(struct hl_ctx *ctx, u64 hop_addr) +{ + struct pgt_info *pgt_info = hl_mmu_dr_get_pgt_info(ctx, hop_addr); + int num_of_ptes_left; + + pgt_info->num_of_ptes--; + + /* + * Need to save the number of ptes left because hl_mmu_free_hop might free + * the pgt_info + */ + num_of_ptes_left = pgt_info->num_of_ptes; + if (!num_of_ptes_left) + hl_mmu_dr_free_pgt_node(ctx, pgt_info); + + return num_of_ptes_left; +} + +u64 hl_mmu_dr_alloc_hop(struct hl_ctx *ctx) +{ + struct hl_device *hdev = ctx->hdev; + struct asic_fixed_properties *prop = &hdev->asic_prop; + struct pgt_info *pgt_info; + u64 phys_addr, shadow_addr; + + pgt_info = kmalloc(sizeof(*pgt_info), GFP_KERNEL); + if (!pgt_info) + return ULLONG_MAX; + + phys_addr = (u64) gen_pool_alloc(hdev->mmu_priv.dr.mmu_pgt_pool, + prop->dmmu.hop_table_size); + if (!phys_addr) { + dev_err(hdev->dev, "failed to allocate page\n"); + goto pool_add_err; + } + + shadow_addr = (u64) (uintptr_t) kzalloc(prop->dmmu.hop_table_size, + GFP_KERNEL); + if (!shadow_addr) + goto shadow_err; + + pgt_info->phys_addr = phys_addr; + pgt_info->shadow_addr = shadow_addr; + pgt_info->ctx = ctx; + pgt_info->num_of_ptes = 0; + hash_add(ctx->mmu_shadow_hash, &pgt_info->node, shadow_addr); + + return shadow_addr; + +shadow_err: + gen_pool_free(hdev->mmu_priv.dr.mmu_pgt_pool, + phys_addr, prop->dmmu.hop_table_size); +pool_add_err: + kfree(pgt_info); + + return ULLONG_MAX; +} + +u64 hl_mmu_dr_get_alloc_next_hop_addr(struct hl_ctx *ctx, u64 curr_pte, bool *is_new_hop) +{ + u64 hop_addr = hl_mmu_get_next_hop_addr(ctx, curr_pte); + + if (hop_addr == ULLONG_MAX) { + hop_addr = hl_mmu_dr_alloc_hop(ctx); + *is_new_hop = (hop_addr != ULLONG_MAX); + } + + return hop_addr; +} + +void hl_mmu_dr_flush(struct hl_ctx *ctx) +{ + /* flush all writes from all cores to reach PCI */ + mb(); + ctx->hdev->asic_funcs->read_pte(ctx->hdev, hl_mmu_dr_get_phys_hop0_addr(ctx)); +} + +int hl_mmu_dr_init(struct hl_device *hdev) +{ + struct asic_fixed_properties *prop = &hdev->asic_prop; + int rc; + + hdev->mmu_priv.dr.mmu_pgt_pool = + gen_pool_create(__ffs(prop->dmmu.hop_table_size), -1); + + if (!hdev->mmu_priv.dr.mmu_pgt_pool) { + dev_err(hdev->dev, "Failed to create page gen pool\n"); + return -ENOMEM; + } + + rc = gen_pool_add(hdev->mmu_priv.dr.mmu_pgt_pool, prop->mmu_pgt_addr + + prop->dmmu.hop0_tables_total_size, + prop->dmmu.pgt_size - prop->dmmu.hop0_tables_total_size, + -1); + if (rc) { + dev_err(hdev->dev, "Failed to add memory to page gen pool\n"); + goto err_pool_add; + } + + hdev->mmu_priv.dr.mmu_shadow_hop0 = kvcalloc(prop->max_asid, + prop->dmmu.hop_table_size, GFP_KERNEL); + if (ZERO_OR_NULL_PTR(hdev->mmu_priv.dr.mmu_shadow_hop0)) { + rc = -ENOMEM; + goto err_pool_add; + } + + /* MMU H/W init will be done in device hw_init() */ + + return 0; + +err_pool_add: + gen_pool_destroy(hdev->mmu_priv.dr.mmu_pgt_pool); + + return rc; +} + +void hl_mmu_dr_fini(struct hl_device *hdev) +{ + /* MMU H/W fini was already done in device hw_fini() */ + + if (ZERO_OR_NULL_PTR(hdev->mmu_priv.dr.mmu_shadow_hop0)) + return; + + kvfree(hdev->mmu_priv.dr.mmu_shadow_hop0); + gen_pool_destroy(hdev->mmu_priv.dr.mmu_pgt_pool); + + /* Make sure that if we arrive here again without init was + * called we won't cause kernel panic. This can happen for + * example if we fail during hard reset code at certain points + */ + hdev->mmu_priv.dr.mmu_shadow_hop0 = NULL; +} diff --git a/drivers/accel/habanalabs/common/mmu/mmu_v1.c b/drivers/accel/habanalabs/common/mmu/mmu_v1.c index d925dc4dd097..845d16aaa637 100644 --- a/drivers/accel/habanalabs/common/mmu/mmu_v1.c +++ b/drivers/accel/habanalabs/common/mmu/mmu_v1.c @@ -12,166 +12,6 @@ #define MMU_V1_MAX_HOPS (MMU_HOP4 + 1) -static inline u64 get_phys_addr(struct hl_ctx *ctx, u64 shadow_addr); - -static struct pgt_info *get_pgt_info(struct hl_ctx *ctx, u64 hop_addr) -{ - struct pgt_info *pgt_info = NULL; - - hash_for_each_possible(ctx->mmu_shadow_hash, pgt_info, node, - (unsigned long) hop_addr) - if (hop_addr == pgt_info->shadow_addr) - break; - - return pgt_info; -} - -static void _free_hop(struct hl_ctx *ctx, struct pgt_info *pgt_info) -{ - struct hl_device *hdev = ctx->hdev; - - gen_pool_free(hdev->mmu_priv.dr.mmu_pgt_pool, pgt_info->phys_addr, - hdev->asic_prop.mmu_hop_table_size); - hash_del(&pgt_info->node); - kfree((u64 *) (uintptr_t) pgt_info->shadow_addr); - kfree(pgt_info); -} - -static void free_hop(struct hl_ctx *ctx, u64 hop_addr) -{ - struct pgt_info *pgt_info = get_pgt_info(ctx, hop_addr); - - _free_hop(ctx, pgt_info); -} - -static u64 alloc_hop(struct hl_ctx *ctx) -{ - struct hl_device *hdev = ctx->hdev; - struct asic_fixed_properties *prop = &hdev->asic_prop; - struct pgt_info *pgt_info; - u64 phys_addr, shadow_addr; - - pgt_info = kmalloc(sizeof(*pgt_info), GFP_KERNEL); - if (!pgt_info) - return ULLONG_MAX; - - phys_addr = (u64) gen_pool_alloc(hdev->mmu_priv.dr.mmu_pgt_pool, - prop->mmu_hop_table_size); - if (!phys_addr) { - dev_err(hdev->dev, "failed to allocate page\n"); - goto pool_add_err; - } - - shadow_addr = (u64) (uintptr_t) kzalloc(prop->mmu_hop_table_size, - GFP_KERNEL); - if (!shadow_addr) - goto shadow_err; - - pgt_info->phys_addr = phys_addr; - pgt_info->shadow_addr = shadow_addr; - pgt_info->ctx = ctx; - pgt_info->num_of_ptes = 0; - hash_add(ctx->mmu_shadow_hash, &pgt_info->node, shadow_addr); - - return shadow_addr; - -shadow_err: - gen_pool_free(hdev->mmu_priv.dr.mmu_pgt_pool, phys_addr, - prop->mmu_hop_table_size); -pool_add_err: - kfree(pgt_info); - - return ULLONG_MAX; -} - -static inline u64 get_phys_hop0_addr(struct hl_ctx *ctx) -{ - return ctx->hdev->asic_prop.mmu_pgt_addr + - (ctx->asid * ctx->hdev->asic_prop.mmu_hop_table_size); -} - -static inline u64 get_hop0_addr(struct hl_ctx *ctx) -{ - return (u64) (uintptr_t) ctx->hdev->mmu_priv.dr.mmu_shadow_hop0 + - (ctx->asid * ctx->hdev->asic_prop.mmu_hop_table_size); -} - -static void flush(struct hl_ctx *ctx) -{ - /* flush all writes from all cores to reach PCI */ - mb(); - ctx->hdev->asic_funcs->read_pte(ctx->hdev, get_phys_hop0_addr(ctx)); -} - -/* transform the value to physical address when writing to H/W */ -static inline void write_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, u64 val) -{ - /* - * The value to write is actually the address of the next shadow hop + - * flags at the 12 LSBs. - * Hence in order to get the value to write to the physical PTE, we - * clear the 12 LSBs and translate the shadow hop to its associated - * physical hop, and add back the original 12 LSBs. - */ - u64 phys_val = get_phys_addr(ctx, val & HOP_PHYS_ADDR_MASK) | - (val & FLAGS_MASK); - - ctx->hdev->asic_funcs->write_pte(ctx->hdev, - get_phys_addr(ctx, shadow_pte_addr), - phys_val); - - *(u64 *) (uintptr_t) shadow_pte_addr = val; -} - -/* do not transform the value to physical address when writing to H/W */ -static inline void write_final_pte(struct hl_ctx *ctx, u64 shadow_pte_addr, - u64 val) -{ - ctx->hdev->asic_funcs->write_pte(ctx->hdev, - get_phys_addr(ctx, shadow_pte_addr), - val); - *(u64 *) (uintptr_t) shadow_pte_addr = val; -} - -/* clear the last and present bits */ -static inline void clear_pte(struct hl_ctx *ctx, u64 pte_addr) -{ - /* no need to transform the value to physical address */ - write_final_pte(ctx, pte_addr, 0); -} - -static inline void get_pte(struct hl_ctx *ctx, u64 hop_addr) -{ - get_pgt_info(ctx, hop_addr)->num_of_ptes++; -} - -/* - * put_pte - decrement the num of ptes and free the hop if possible - * - * @ctx: pointer to the context structure - * @hop_addr: addr of the hop - * - * This function returns the number of ptes left on this hop. If the number is - * 0, it means the pte was freed. - */ -static inline int put_pte(struct hl_ctx *ctx, u64 hop_addr) -{ - struct pgt_info *pgt_info = get_pgt_info(ctx, hop_addr); - int num_of_ptes_left; - - pgt_info->num_of_ptes--; - - /* - * Need to save the number of ptes left because free_hop might free - * the pgt_info - */ - num_of_ptes_left = pgt_info->num_of_ptes; - if (!num_of_ptes_left) - _free_hop(ctx, pgt_info); - - return num_of_ptes_left; -} - static inline u64 get_hop_pte_addr(struct hl_ctx *ctx, struct hl_mmu_properties *mmu_prop, u64 *hop_addr_arr, u64 virt_addr, enum mmu_hop_num hop_idx) { @@ -183,35 +23,6 @@ static inline u64 get_hop_pte_addr(struct hl_ctx *ctx, struct hl_mmu_properties ctx->hdev->asic_prop.mmu_pte_size * ((virt_addr & mask) >> shift); } -static inline u64 get_alloc_next_hop_addr(struct hl_ctx *ctx, u64 curr_pte, - bool *is_new_hop) -{ - u64 hop_addr = hl_mmu_get_next_hop_addr(ctx, curr_pte); - - if (hop_addr == ULLONG_MAX) { - hop_addr = alloc_hop(ctx); - *is_new_hop = (hop_addr != ULLONG_MAX); - } - - return hop_addr; -} - -/* translates shadow address inside hop to a physical address */ -static inline u64 get_phys_addr(struct hl_ctx *ctx, u64 shadow_addr) -{ - u64 page_mask = (ctx->hdev->asic_prop.mmu_hop_table_size - 1); - u64 shadow_hop_addr = shadow_addr & ~page_mask; - u64 pte_offset = shadow_addr & page_mask; - u64 phys_hop_addr; - - if (shadow_hop_addr != get_hop0_addr(ctx)) - phys_hop_addr = get_pgt_info(ctx, shadow_hop_addr)->phys_addr; - else - phys_hop_addr = get_phys_hop0_addr(ctx); - - return phys_hop_addr + pte_offset; -} - static int dram_default_mapping_init(struct hl_ctx *ctx) { struct hl_device *hdev = ctx->hdev; @@ -232,13 +43,13 @@ static int dram_default_mapping_init(struct hl_ctx *ctx) /* add hop1 and hop2 */ total_hops = num_of_hop3 + 2; - ctx->dram_default_hops = kzalloc(HL_PTE_SIZE * total_hops, GFP_KERNEL); + ctx->dram_default_hops = kcalloc(total_hops, HL_PTE_SIZE, GFP_KERNEL); if (!ctx->dram_default_hops) return -ENOMEM; - hop0_addr = get_hop0_addr(ctx); + hop0_addr = hl_mmu_dr_get_hop0_addr(ctx); - hop1_addr = alloc_hop(ctx); + hop1_addr = hl_mmu_dr_alloc_hop(ctx); if (hop1_addr == ULLONG_MAX) { dev_err(hdev->dev, "failed to alloc hop 1\n"); rc = -ENOMEM; @@ -247,7 +58,7 @@ static int dram_default_mapping_init(struct hl_ctx *ctx) ctx->dram_default_hops[total_hops - 1] = hop1_addr; - hop2_addr = alloc_hop(ctx); + hop2_addr = hl_mmu_dr_alloc_hop(ctx); if (hop2_addr == ULLONG_MAX) { dev_err(hdev->dev, "failed to alloc hop 2\n"); rc = -ENOMEM; @@ -257,7 +68,7 @@ static int dram_default_mapping_init(struct hl_ctx *ctx) ctx->dram_default_hops[total_hops - 2] = hop2_addr; for (i = 0 ; i < num_of_hop3 ; i++) { - ctx->dram_default_hops[i] = alloc_hop(ctx); + ctx->dram_default_hops[i] = hl_mmu_dr_alloc_hop(ctx); if (ctx->dram_default_hops[i] == ULLONG_MAX) { dev_err(hdev->dev, "failed to alloc hop 3, i: %d\n", i); rc = -ENOMEM; @@ -268,18 +79,18 @@ static int dram_default_mapping_init(struct hl_ctx *ctx) /* need only pte 0 in hops 0 and 1 */ pte_val = (hop1_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK; - write_pte(ctx, hop0_addr, pte_val); + hl_mmu_dr_write_pte(ctx, hop0_addr, pte_val); pte_val = (hop2_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK; - write_pte(ctx, hop1_addr, pte_val); - get_pte(ctx, hop1_addr); + hl_mmu_dr_write_pte(ctx, hop1_addr, pte_val); + hl_mmu_dr_get_pte(ctx, hop1_addr); hop2_pte_addr = hop2_addr; for (i = 0 ; i < num_of_hop3 ; i++) { pte_val = (ctx->dram_default_hops[i] & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK; - write_pte(ctx, hop2_pte_addr, pte_val); - get_pte(ctx, hop2_addr); + hl_mmu_dr_write_pte(ctx, hop2_pte_addr, pte_val); + hl_mmu_dr_get_pte(ctx, hop2_addr); hop2_pte_addr += HL_PTE_SIZE; } @@ -289,23 +100,23 @@ static int dram_default_mapping_init(struct hl_ctx *ctx) for (i = 0 ; i < num_of_hop3 ; i++) { hop3_pte_addr = ctx->dram_default_hops[i]; for (j = 0 ; j < HOP_PTE_ENTRIES_512 ; j++) { - write_final_pte(ctx, hop3_pte_addr, pte_val); - get_pte(ctx, ctx->dram_default_hops[i]); + hl_mmu_dr_write_final_pte(ctx, hop3_pte_addr, pte_val); + hl_mmu_dr_get_pte(ctx, ctx->dram_default_hops[i]); hop3_pte_addr += HL_PTE_SIZE; } } - flush(ctx); + hl_mmu_dr_flush(ctx); return 0; hop3_err: for (i = 0 ; i < hop3_allocated ; i++) - free_hop(ctx, ctx->dram_default_hops[i]); + hl_mmu_dr_free_hop(ctx, ctx->dram_default_hops[i]); - free_hop(ctx, hop2_addr); + hl_mmu_dr_free_hop(ctx, hop2_addr); hop2_err: - free_hop(ctx, hop1_addr); + hl_mmu_dr_free_hop(ctx, hop1_addr); hop1_err: kfree(ctx->dram_default_hops); @@ -329,7 +140,7 @@ static void dram_default_mapping_fini(struct hl_ctx *ctx) do_div(num_of_hop3, prop->dram_page_size); do_div(num_of_hop3, HOP_PTE_ENTRIES_512); - hop0_addr = get_hop0_addr(ctx); + hop0_addr = hl_mmu_dr_get_hop0_addr(ctx); /* add hop1 and hop2 */ total_hops = num_of_hop3 + 2; hop1_addr = ctx->dram_default_hops[total_hops - 1]; @@ -338,101 +149,26 @@ static void dram_default_mapping_fini(struct hl_ctx *ctx) for (i = 0 ; i < num_of_hop3 ; i++) { hop3_pte_addr = ctx->dram_default_hops[i]; for (j = 0 ; j < HOP_PTE_ENTRIES_512 ; j++) { - clear_pte(ctx, hop3_pte_addr); - put_pte(ctx, ctx->dram_default_hops[i]); + hl_mmu_dr_clear_pte(ctx, hop3_pte_addr); + hl_mmu_dr_put_pte(ctx, ctx->dram_default_hops[i]); hop3_pte_addr += HL_PTE_SIZE; } } hop2_pte_addr = hop2_addr; for (i = 0 ; i < num_of_hop3 ; i++) { - clear_pte(ctx, hop2_pte_addr); - put_pte(ctx, hop2_addr); + hl_mmu_dr_clear_pte(ctx, hop2_pte_addr); + hl_mmu_dr_put_pte(ctx, hop2_addr); hop2_pte_addr += HL_PTE_SIZE; } - clear_pte(ctx, hop1_addr); - put_pte(ctx, hop1_addr); - clear_pte(ctx, hop0_addr); + hl_mmu_dr_clear_pte(ctx, hop1_addr); + hl_mmu_dr_put_pte(ctx, hop1_addr); + hl_mmu_dr_clear_pte(ctx, hop0_addr); kfree(ctx->dram_default_hops); - flush(ctx); -} - -/** - * hl_mmu_v1_init() - initialize the MMU module. - * @hdev: habanalabs device structure. - * - * This function does the following: - * - Create a pool of pages for pgt_infos. - * - Create a shadow table for pgt - * - * Return: 0 for success, non-zero for failure. - */ -static int hl_mmu_v1_init(struct hl_device *hdev) -{ - struct asic_fixed_properties *prop = &hdev->asic_prop; - int rc; - - hdev->mmu_priv.dr.mmu_pgt_pool = - gen_pool_create(__ffs(prop->mmu_hop_table_size), -1); - - if (!hdev->mmu_priv.dr.mmu_pgt_pool) { - dev_err(hdev->dev, "Failed to create page gen pool\n"); - return -ENOMEM; - } - - rc = gen_pool_add(hdev->mmu_priv.dr.mmu_pgt_pool, prop->mmu_pgt_addr + - prop->mmu_hop0_tables_total_size, - prop->mmu_pgt_size - prop->mmu_hop0_tables_total_size, - -1); - if (rc) { - dev_err(hdev->dev, "Failed to add memory to page gen pool\n"); - goto err_pool_add; - } - - hdev->mmu_priv.dr.mmu_shadow_hop0 = kvcalloc(prop->max_asid, prop->mmu_hop_table_size, - GFP_KERNEL); - if (ZERO_OR_NULL_PTR(hdev->mmu_priv.dr.mmu_shadow_hop0)) { - rc = -ENOMEM; - goto err_pool_add; - } - - /* MMU H/W init will be done in device hw_init() */ - - return 0; - -err_pool_add: - gen_pool_destroy(hdev->mmu_priv.dr.mmu_pgt_pool); - - return rc; -} - -/** - * hl_mmu_v1_fini() - release the MMU module. - * @hdev: habanalabs device structure. - * - * This function does the following: - * - Disable MMU in H/W. - * - Free the pgt_infos pool. - * - * All contexts should be freed before calling this function. - */ -static void hl_mmu_v1_fini(struct hl_device *hdev) -{ - /* MMU H/W fini was already done in device hw_fini() */ - - if (!ZERO_OR_NULL_PTR(hdev->mmu_priv.dr.mmu_shadow_hop0)) { - kvfree(hdev->mmu_priv.dr.mmu_shadow_hop0); - gen_pool_destroy(hdev->mmu_priv.dr.mmu_pgt_pool); - - /* Make sure that if we arrive here again without init was - * called we won't cause kernel panic. This can happen for - * example if we fail during hard reset code at certain points - */ - hdev->mmu_priv.dr.mmu_shadow_hop0 = NULL; - } + hl_mmu_dr_flush(ctx); } /** @@ -476,7 +212,7 @@ static void hl_mmu_v1_ctx_fini(struct hl_ctx *ctx) dev_err_ratelimited(hdev->dev, "pgt_info of addr 0x%llx of asid %d was not destroyed, num_ptes: %d\n", pgt_info->phys_addr, ctx->asid, pgt_info->num_of_ptes); - _free_hop(ctx, pgt_info); + hl_mmu_dr_free_pgt_node(ctx, pgt_info); } } @@ -495,7 +231,7 @@ static int hl_mmu_v1_unmap(struct hl_ctx *ctx, for (hop_idx = MMU_HOP0; hop_idx < MMU_HOP4; hop_idx++) { if (hop_idx == MMU_HOP0) { - hop_addr[hop_idx] = get_hop0_addr(ctx); + hop_addr[hop_idx] = hl_mmu_dr_get_hop0_addr(ctx); } else { hop_addr[hop_idx] = hl_mmu_get_next_hop_addr(ctx, curr_pte); if (hop_addr[hop_idx] == ULLONG_MAX) @@ -546,30 +282,30 @@ static int hl_mmu_v1_unmap(struct hl_ctx *ctx, } hop_idx = MMU_HOP3; - write_final_pte(ctx, hop_pte_addr[hop_idx], default_pte); - put_pte(ctx, hop_addr[hop_idx]); + hl_mmu_dr_write_final_pte(ctx, hop_pte_addr[hop_idx], default_pte); + hl_mmu_dr_put_pte(ctx, hop_addr[hop_idx]); } else { if (!(curr_pte & PAGE_PRESENT_MASK)) goto not_mapped; if (hop_addr[MMU_HOP4]) - clear_pte(ctx, hop_pte_addr[MMU_HOP4]); + hl_mmu_dr_clear_pte(ctx, hop_pte_addr[MMU_HOP4]); else - clear_pte(ctx, hop_pte_addr[MMU_HOP3]); + hl_mmu_dr_clear_pte(ctx, hop_pte_addr[MMU_HOP3]); - if (hop_addr[MMU_HOP4] && !put_pte(ctx, hop_addr[MMU_HOP4])) + if (hop_addr[MMU_HOP4] && !hl_mmu_dr_put_pte(ctx, hop_addr[MMU_HOP4])) clear_hop3 = true; if (!clear_hop3) goto mapped; for (hop_idx = MMU_HOP3; hop_idx >= 0; hop_idx--) { - clear_pte(ctx, hop_pte_addr[hop_idx]); + hl_mmu_dr_clear_pte(ctx, hop_pte_addr[hop_idx]); if (hop_idx == MMU_HOP0) break; - if (put_pte(ctx, hop_addr[hop_idx])) + if (hl_mmu_dr_put_pte(ctx, hop_addr[hop_idx])) goto mapped; } } @@ -616,10 +352,10 @@ static int hl_mmu_v1_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, for (hop_idx = MMU_HOP0; hop_idx < num_hops; hop_idx++) { if (hop_idx == MMU_HOP0) { - hop_addr[hop_idx] = get_hop0_addr(ctx); + hop_addr[hop_idx] = hl_mmu_dr_get_hop0_addr(ctx); } else { hop_addr[hop_idx] = - get_alloc_next_hop_addr(ctx, curr_pte, &hop_new[hop_idx]); + hl_mmu_dr_get_alloc_next_hop_addr(ctx, curr_pte, &hop_new[hop_idx]); if (hop_addr[hop_idx] == ULLONG_MAX) goto err; } @@ -666,27 +402,27 @@ static int hl_mmu_v1_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, curr_pte = (phys_addr & HOP_PHYS_ADDR_MASK) | mmu_prop->last_mask | PAGE_PRESENT_MASK; - write_final_pte(ctx, hop_pte_addr[num_hops - 1], curr_pte); + hl_mmu_dr_write_final_pte(ctx, hop_pte_addr[num_hops - 1], curr_pte); for (hop_idx = MMU_HOP1; hop_idx < num_hops; hop_idx++) { prev_hop = hop_idx - 1; if (hop_new[hop_idx]) { curr_pte = (hop_addr[hop_idx] & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK; - write_pte(ctx, hop_pte_addr[prev_hop], curr_pte); + hl_mmu_dr_write_pte(ctx, hop_pte_addr[prev_hop], curr_pte); if (hop_idx != MMU_HOP1) - get_pte(ctx, hop_addr[prev_hop]); + hl_mmu_dr_get_pte(ctx, hop_addr[prev_hop]); } } - get_pte(ctx, hop_addr[num_hops - 1]); + hl_mmu_dr_get_pte(ctx, hop_addr[num_hops - 1]); return 0; err: for (hop_idx = num_hops; hop_idx > MMU_HOP0; hop_idx--) { if (hop_new[hop_idx]) - free_hop(ctx, hop_addr[hop_idx]); + hl_mmu_dr_free_hop(ctx, hop_addr[hop_idx]); } return rc; @@ -752,7 +488,7 @@ static int hl_mmu_v1_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, if (is_huge) used_hops--; - hops->hop_info[0].hop_addr = get_phys_hop0_addr(ctx); + hops->hop_info[0].hop_addr = hl_mmu_dr_get_phys_hop0_addr(ctx); hops->hop_info[0].hop_pte_addr = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, 0, hops->hop_info[0].hop_addr, virt_addr); @@ -801,13 +537,13 @@ static int hl_mmu_v1_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, */ void hl_mmu_v1_set_funcs(struct hl_device *hdev, struct hl_mmu_funcs *mmu) { - mmu->init = hl_mmu_v1_init; - mmu->fini = hl_mmu_v1_fini; + mmu->init = hl_mmu_dr_init; + mmu->fini = hl_mmu_dr_fini; mmu->ctx_init = hl_mmu_v1_ctx_init; mmu->ctx_fini = hl_mmu_v1_ctx_fini; mmu->map = hl_mmu_v1_map; mmu->unmap = hl_mmu_v1_unmap; - mmu->flush = flush; + mmu->flush = hl_mmu_dr_flush; mmu->swap_out = hl_mmu_v1_swap_out; mmu->swap_in = hl_mmu_v1_swap_in; mmu->get_tlb_info = hl_mmu_v1_get_tlb_info; diff --git a/drivers/accel/habanalabs/common/mmu/mmu_v2.c b/drivers/accel/habanalabs/common/mmu/mmu_v2.c new file mode 100644 index 000000000000..4bc0268fff1c --- /dev/null +++ b/drivers/accel/habanalabs/common/mmu/mmu_v2.c @@ -0,0 +1,338 @@ +// SPDX-License-Identifier: GPL-2.0 + +/* + * Copyright 2016-2020 HabanaLabs, Ltd. + * All Rights Reserved. + */ + +#include "../habanalabs.h" +#include "../../include/hw_ip/mmu/mmu_general.h" +#include "../../include/hw_ip/mmu/mmu_v2_0.h" + +#include <linux/slab.h> + +/** + * hl_mmu_v2_ctx_init() - initialize a context for using the MMU module. + * @ctx: pointer to the context structure to initialize. + * + * Initialize a mutex to protect the concurrent mapping flow, a hash to hold all + * page tables hops related to this context. + * Return: 0 on success, non-zero otherwise. + */ +static int hl_mmu_v2_ctx_init(struct hl_ctx *ctx) +{ + hash_init(ctx->mmu_shadow_hash); + + return 0; +} + +/* + * hl_mmu_v2_ctx_fini - disable a ctx from using the mmu module + * + * @ctx: pointer to the context structure + * + * This function does the following: + * - Free any pgts which were not freed yet + * - Free the mutex + * - Free DRAM default page mapping hops + */ +static void hl_mmu_v2_ctx_fini(struct hl_ctx *ctx) +{ + struct hl_device *hdev = ctx->hdev; + struct pgt_info *pgt_info; + struct hlist_node *tmp; + int i; + + if (!hash_empty(ctx->mmu_shadow_hash)) + dev_err(hdev->dev, "ctx %d is freed while it has pgts in use\n", + ctx->asid); + + hash_for_each_safe(ctx->mmu_shadow_hash, i, tmp, pgt_info, node) { + dev_err_ratelimited(hdev->dev, + "pgt_info of addr 0x%llx of asid %d was not destroyed, num_ptes: %d\n", + pgt_info->phys_addr, ctx->asid, pgt_info->num_of_ptes); + hl_mmu_dr_free_pgt_node(ctx, pgt_info); + } +} + +static int hl_mmu_v2_unmap(struct hl_ctx *ctx, u64 virt_addr, bool is_dram_addr) +{ + u64 hop_addr[MMU_ARCH_6_HOPS] = { 0 }, hop_pte_addr[MMU_ARCH_6_HOPS] = { 0 }, curr_pte, + scrambled_virt_addr; + struct asic_fixed_properties *prop = &ctx->hdev->asic_prop; + struct hl_device *hdev = ctx->hdev; + struct hl_mmu_properties *mmu_prop; + bool is_huge = false; + int i, hop_last; + + /* device resident in V2 are allowed only for HMMU */ + if (!is_dram_addr) + return -EINVAL; + + mmu_prop = &prop->dmmu; + + hop_last = mmu_prop->num_hops - 1; + + scrambled_virt_addr = hdev->asic_funcs->scramble_addr(hdev, virt_addr); + + hop_addr[0] = hl_mmu_dr_get_hop0_addr(ctx); + hop_pte_addr[0] = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, 0, + hop_addr[0], scrambled_virt_addr); + if (hop_pte_addr[0] == U64_MAX) + return -EFAULT; + + curr_pte = *(u64 *) (uintptr_t) hop_pte_addr[0]; + + for (i = 1 ; i < mmu_prop->num_hops ; i++) { + hop_addr[i] = hl_mmu_get_next_hop_addr(ctx, curr_pte); + if (hop_addr[i] == ULLONG_MAX) + goto not_mapped; + + hop_pte_addr[i] = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, i, + hop_addr[i], scrambled_virt_addr); + if (hop_pte_addr[i] == U64_MAX) + return -EFAULT; + + curr_pte = *(u64 *) (uintptr_t) hop_pte_addr[i]; + + if ((i <= hop_last) && (curr_pte & mmu_prop->last_mask)) { + hop_last = i; + is_huge = true; + break; + } + } + + if (is_dram_addr && !is_huge) { + dev_err(hdev->dev, "DRAM unmapping should use huge pages only\n"); + return -EFAULT; + } + + if (!(curr_pte & PAGE_PRESENT_MASK)) + goto not_mapped; + + for (i = hop_last ; i > 0 ; i--) { + hl_mmu_dr_clear_pte(ctx, hop_pte_addr[i]); + if (hl_mmu_dr_put_pte(ctx, hop_addr[i])) + goto mapped; + } + hl_mmu_dr_clear_pte(ctx, hop_pte_addr[0]); + +mapped: + return 0; + +not_mapped: + dev_err(hdev->dev, "virt addr 0x%llx is not mapped to phys addr\n", + virt_addr); + + return -EINVAL; +} + +static int hl_mmu_v2_map(struct hl_ctx *ctx, u64 virt_addr, u64 phys_addr, + u32 page_size, bool is_dram_addr) +{ + u64 hop_addr[MMU_ARCH_6_HOPS] = { 0 }, hop_pte_addr[MMU_ARCH_6_HOPS] = { 0 }, + curr_pte = 0, scrambled_virt_addr, scrambled_phys_addr; + struct asic_fixed_properties *prop = &ctx->hdev->asic_prop; + bool hop_new[MMU_ARCH_6_HOPS] = { false }; + struct hl_device *hdev = ctx->hdev; + struct hl_mmu_properties *mmu_prop; + int rc, i, hop_last; + + /* device resident in V2 are allowed only for HMMU */ + if (!is_dram_addr) + return -EINVAL; + + mmu_prop = &prop->dmmu; + + hop_last = mmu_prop->num_hops - 1; + + scrambled_virt_addr = hdev->asic_funcs->scramble_addr(hdev, virt_addr); + scrambled_phys_addr = hdev->asic_funcs->scramble_addr(hdev, phys_addr); + + /* First hop is preallocated therefore it is treated differently */ + hop_addr[0] = hl_mmu_dr_get_hop0_addr(ctx); + hop_pte_addr[0] = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, 0, + hop_addr[0], scrambled_virt_addr); + curr_pte = *(u64 *) (uintptr_t) hop_pte_addr[0]; + + /* Handle hop1 to hop_last */ + for (i = 1 ; i <= hop_last ; i++) { + hop_addr[i] = hl_mmu_dr_get_alloc_next_hop_addr(ctx, curr_pte, &hop_new[i]); + if (hop_addr[i] == ULLONG_MAX) { + rc = -ENOMEM; + goto err; + } + + hop_pte_addr[i] = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, i, + hop_addr[i], scrambled_virt_addr); + if (hop_pte_addr[i] == U64_MAX) { + rc = -EINVAL; + goto err; + } + + if (!hop_pte_addr[i]) { + rc = -EINVAL; + goto err; + } + + curr_pte = *(u64 *) (uintptr_t) hop_pte_addr[i]; + } + + if (curr_pte & PAGE_PRESENT_MASK) { + dev_err(hdev->dev, + "mapping already exists for virt_addr 0x%llx\n", + virt_addr); + + for (i = 0 ; i <= hop_last ; i++) + dev_dbg(hdev->dev, "hop%d pte: 0x%llx (0x%llx)\n", + i, *(u64 *) (uintptr_t) hop_pte_addr[i], + hop_pte_addr[i]); + + rc = -EINVAL; + goto err; + } + + curr_pte = (scrambled_phys_addr & HOP_PHYS_ADDR_MASK) + | mmu_prop->last_mask | PAGE_PRESENT_MASK; + + /* Write the PTEs */ + hl_mmu_dr_write_final_pte(ctx, hop_pte_addr[hop_last], curr_pte); + + /* for each new hop, add its address to the table of previous-hop */ + for (i = 1 ; i <= hop_last ; i++) { + if (hop_new[i]) { + curr_pte = (hop_addr[i] & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK; + hl_mmu_dr_write_pte(ctx, hop_pte_addr[i - 1], curr_pte); + + if (i - 1) + hl_mmu_dr_get_pte(ctx, hop_addr[i - 1]); + } + } + hl_mmu_dr_get_pte(ctx, hop_addr[hop_last]); + + return 0; + +err: + for (i = 1 ; i <= hop_last ; i++) + if (hop_new[i] && (hop_addr[i] != U64_MAX)) + hl_mmu_dr_free_hop(ctx, hop_addr[i]); + + return rc; +} + +/* + * hl_mmu_v2_swap_out - marks all mapping of the given ctx as swapped out + * + * @ctx: pointer to the context structure + * + */ +static void hl_mmu_v2_swap_out(struct hl_ctx *ctx) +{ + +} + +/* + * hl_mmu_v2_swap_in - marks all mapping of the given ctx as swapped in + * + * @ctx: pointer to the context structure + * + */ +static void hl_mmu_v2_swap_in(struct hl_ctx *ctx) +{ + +} + +static int hl_mmu_v2_get_tlb_info(struct hl_ctx *ctx, u64 virt_addr, struct hl_mmu_hop_info *hops) +{ + struct asic_fixed_properties *prop = &ctx->hdev->asic_prop; + struct hl_device *hdev = ctx->hdev; + struct hl_mmu_properties *mmu_prop; + bool is_dram_addr; + int i; + + is_dram_addr = hl_mem_area_inside_range(virt_addr, prop->dmmu.page_size, + prop->dmmu.start_addr, + prop->dmmu.end_addr); + + /* device resident in V2 are allowed only for HMMU */ + if (!is_dram_addr) + return -EINVAL; + + mmu_prop = &prop->dmmu; + hops->range_type = HL_VA_RANGE_TYPE_DRAM; + + hops->scrambled_vaddr = hdev->asic_funcs->scramble_addr(hdev, virt_addr); + + hops->hop_info[0].hop_addr = hl_mmu_dr_get_phys_hop0_addr(ctx); + hops->hop_info[0].hop_pte_addr = hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, 0, + hops->hop_info[0].hop_addr, + hops->scrambled_vaddr); + if (hops->hop_info[0].hop_pte_addr == U64_MAX) + return -EFAULT; + + hops->hop_info[0].hop_pte_val = hdev->asic_funcs->read_pte(hdev, + hops->hop_info[0].hop_pte_addr); + if (hops->hop_info[0].hop_pte_val == U64_MAX) + return -EFAULT; + + for (i = 1 ; i < mmu_prop->num_hops ; i++) { + hops->hop_info[i].hop_addr = + hl_mmu_get_next_hop_addr(ctx, hops->hop_info[i - 1].hop_pte_val); + if (hops->hop_info[i].hop_addr == ULLONG_MAX) + return -EFAULT; + + hops->hop_info[i].hop_pte_addr = + hl_mmu_get_hop_pte_phys_addr(ctx, mmu_prop, i, + hops->hop_info[i].hop_addr, + hops->scrambled_vaddr); + if (hops->hop_info[i].hop_pte_addr == U64_MAX) + return -EFAULT; + + hops->hop_info[i].hop_pte_val = + hdev->asic_funcs->read_pte(hdev, + hops->hop_info[i].hop_pte_addr); + + if (!(hops->hop_info[i].hop_pte_val & PAGE_PRESENT_MASK)) + return -EFAULT; + + if (hops->hop_info[i].hop_pte_val & mmu_prop->last_mask) + break; + } + + /* if passed over all hops then no last hop was found */ + if (i == mmu_prop->num_hops) + return -EFAULT; + + if (!(hops->hop_info[i].hop_pte_val & PAGE_PRESENT_MASK)) + return -EFAULT; + + if (hops->scrambled_vaddr != virt_addr) + hops->unscrambled_paddr = hdev->asic_funcs->descramble_addr + (hdev, hops->hop_info[i].hop_pte_val); + else + hops->unscrambled_paddr = hops->hop_info[i].hop_pte_val; + + hops->used_hops = i + 1; + + return 0; +} + +/* + * hl_mmu_v2_prepare - prepare mmu_if for working with mmu v2 + * + * @hdev: pointer to the device structure + * @mmu_if: pointer to the mmu interface structure + */ +void hl_mmu_v2_set_funcs(struct hl_device *hdev, struct hl_mmu_funcs *mmu) +{ + mmu->init = hl_mmu_dr_init; + mmu->fini = hl_mmu_dr_fini; + mmu->ctx_init = hl_mmu_v2_ctx_init; + mmu->ctx_fini = hl_mmu_v2_ctx_fini; + mmu->map = hl_mmu_v2_map; + mmu->unmap = hl_mmu_v2_unmap; + mmu->flush = hl_mmu_dr_flush; + mmu->swap_out = hl_mmu_v2_swap_out; + mmu->swap_in = hl_mmu_v2_swap_in; + mmu->get_tlb_info = hl_mmu_v2_get_tlb_info; +} diff --git a/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c b/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c index afe7ef964f82..31507b2a431b 100644 --- a/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c +++ b/drivers/accel/habanalabs/common/mmu/mmu_v2_hr.c @@ -47,7 +47,7 @@ static inline int hl_mmu_v2_hr_init(struct hl_device *hdev) { struct asic_fixed_properties *prop = &hdev->asic_prop; - return hl_mmu_hr_init(hdev, &hdev->mmu_priv.hr, prop->mmu_hop_table_size, + return hl_mmu_hr_init(hdev, &hdev->mmu_priv.hr, prop->pmmu.hop_table_size, prop->mmu_pgt_size); } @@ -65,7 +65,7 @@ static inline void hl_mmu_v2_hr_fini(struct hl_device *hdev) { struct asic_fixed_properties *prop = &hdev->asic_prop; - hl_mmu_hr_fini(hdev, &hdev->mmu_priv.hr, prop->mmu_hop_table_size); + hl_mmu_hr_fini(hdev, &hdev->mmu_priv.hr, prop->pmmu.hop_table_size); } /** @@ -108,7 +108,7 @@ static void hl_mmu_v2_hr_ctx_fini(struct hl_ctx *ctx) "pgt_info of addr 0x%llx of asid %d was not destroyed, num_ptes: %d\n", pgt_info->phys_addr, ctx->asid, pgt_info->num_of_ptes); hl_mmu_hr_free_hop_remove_pgt(pgt_info, &ctx->hdev->mmu_priv.hr, - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); } } @@ -150,7 +150,7 @@ static int _hl_mmu_v2_hr_unmap(struct hl_ctx *ctx, curr_pte = *(u64 *) (uintptr_t) hl_mmu_hr_pte_phys_to_virt(ctx, hops_pgt_info[i], hop_pte_phys_addr[i], - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); if ((i < hop_last) && (curr_pte & mmu_prop->last_mask)) { hop_last = i; @@ -169,14 +169,14 @@ static int _hl_mmu_v2_hr_unmap(struct hl_ctx *ctx, for (i = hop_last ; i > 0 ; i--) { hl_mmu_hr_clear_pte(ctx, hops_pgt_info[i], hop_pte_phys_addr[i], - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); if (hl_mmu_hr_put_pte(ctx, hops_pgt_info[i], &ctx->hdev->mmu_priv.hr, - ctx->hdev->asic_prop.mmu_hop_table_size)) + ctx->hdev->asic_prop.pmmu.hop_table_size)) goto mapped; } hl_mmu_hr_clear_pte(ctx, hops_pgt_info[0], hop_pte_phys_addr[0], - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); mapped: return 0; @@ -255,7 +255,7 @@ static int _hl_mmu_v2_hr_map(struct hl_ctx *ctx, scrambled_virt_addr); curr_pte = *(u64 *) (uintptr_t) hl_mmu_hr_pte_phys_to_virt(ctx, hops_pgt_info[i], hop_pte_phys_addr[i], - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); } if (curr_pte & PAGE_PRESENT_MASK) { @@ -268,7 +268,7 @@ static int _hl_mmu_v2_hr_map(struct hl_ctx *ctx, *(u64 *) (uintptr_t) hl_mmu_hr_pte_phys_to_virt(ctx, hops_pgt_info[i], hop_pte_phys_addr[i], - ctx->hdev->asic_prop.mmu_hop_table_size), + ctx->hdev->asic_prop.pmmu.hop_table_size), hop_pte_phys_addr[i]); rc = -EINVAL; goto err; @@ -279,7 +279,7 @@ static int _hl_mmu_v2_hr_map(struct hl_ctx *ctx, /* Write the PTEs */ hl_mmu_hr_write_pte(ctx, hops_pgt_info[hop_last], hop_pte_phys_addr[hop_last], curr_pte, - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); /* for each new hop, add its address to the table of previous-hop */ for (i = 1 ; i <= hop_last ; i++) { @@ -287,7 +287,7 @@ static int _hl_mmu_v2_hr_map(struct hl_ctx *ctx, curr_pte = (hops_pgt_info[i]->phys_addr & HOP_PHYS_ADDR_MASK) | PAGE_PRESENT_MASK; hl_mmu_hr_write_pte(ctx, hops_pgt_info[i - 1], hop_pte_phys_addr[i - 1], - curr_pte, ctx->hdev->asic_prop.mmu_hop_table_size); + curr_pte, ctx->hdev->asic_prop.pmmu.hop_table_size); if (i - 1) hl_mmu_hr_get_pte(ctx, &ctx->hdev->mmu_func[MMU_HR_PGT].hr_funcs, hops_pgt_info[i - 1]->phys_addr); @@ -303,7 +303,7 @@ err: for (i = 1 ; i <= hop_last ; i++) if (hop_new[i] && hops_pgt_info[i]) hl_mmu_hr_free_hop_remove_pgt(hops_pgt_info[i], &ctx->hdev->mmu_priv.hr, - ctx->hdev->asic_prop.mmu_hop_table_size); + ctx->hdev->asic_prop.pmmu.hop_table_size); return rc; } diff --git a/drivers/accel/habanalabs/common/pci/pci.c b/drivers/accel/habanalabs/common/pci/pci.c index 191e0e3cf3a5..81cbd8697d4c 100644 --- a/drivers/accel/habanalabs/common/pci/pci.c +++ b/drivers/accel/habanalabs/common/pci/pci.c @@ -123,7 +123,7 @@ int hl_pci_elbi_read(struct hl_device *hdev, u64 addr, u32 *data) pci_read_config_dword(pdev, mmPCI_CONFIG_ELBI_DATA, data); if (unlikely(trace_habanalabs_elbi_read_enabled())) - trace_habanalabs_elbi_read(hdev->dev, (u32) addr, val); + trace_habanalabs_elbi_read(&hdev->pdev->dev, (u32) addr, val); return 0; } @@ -186,7 +186,7 @@ static int hl_pci_elbi_write(struct hl_device *hdev, u64 addr, u32 data) if ((val & PCI_CONFIG_ELBI_STS_MASK) == PCI_CONFIG_ELBI_STS_DONE) { if (unlikely(trace_habanalabs_elbi_write_enabled())) - trace_habanalabs_elbi_write(hdev->dev, (u32) addr, val); + trace_habanalabs_elbi_write(&hdev->pdev->dev, (u32) addr, val); return 0; } diff --git a/drivers/accel/habanalabs/common/security.c b/drivers/accel/habanalabs/common/security.c index fe913965dbad..5402a3cd0491 100644 --- a/drivers/accel/habanalabs/common/security.c +++ b/drivers/accel/habanalabs/common/security.c @@ -7,15 +7,31 @@ #include "habanalabs.h" -static const char * const hl_glbl_error_cause[HL_MAX_NUM_OF_GLBL_ERR_CAUSE] = { +static const char * const hl_glbl_error_cause[] = { "Error due to un-priv read", "Error due to un-secure read", "Error due to read from unmapped reg", "Error due to un-priv write", "Error due to un-secure write", "Error due to write to unmapped reg", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", "External I/F write sec violation", "External I/F write to un-mapped reg", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", + "N/A", "Read to write only", "Write to read only" }; @@ -671,10 +687,11 @@ static bool hl_check_block_range_exclusion(struct hl_device *hdev, static int hl_read_glbl_errors(struct hl_device *hdev, u32 blk_idx, u32 major, u32 minor, u32 sub_minor, void *data) { - struct hl_special_block_info *special_blocks = hdev->asic_prop.special_blocks; + struct asic_fixed_properties *prop = &hdev->asic_prop; + struct hl_special_block_info *special_blocks = prop->special_blocks; struct hl_special_block_info *current_block = &special_blocks[blk_idx]; u32 glbl_err_addr, glbl_err_cause, addr_val, cause_val, block_base, - base = current_block->base_addr - lower_32_bits(hdev->asic_prop.cfg_base_address); + base = current_block->base_addr - lower_32_bits(prop->cfg_base_address); int i; block_base = base + major * current_block->major_offset + @@ -689,13 +706,13 @@ static int hl_read_glbl_errors(struct hl_device *hdev, glbl_err_addr = block_base + HL_GLBL_ERR_ADDR_OFFSET; addr_val = RREG32(glbl_err_addr); - for (i = 0 ; i < hdev->asic_prop.glbl_err_cause_num ; i++) { + for (i = 0 ; i <= prop->glbl_err_max_cause_num ; i++) { if (cause_val & BIT(i)) dev_err_ratelimited(hdev->dev, - "%s, addr %#llx\n", - hl_glbl_error_cause[i], - hdev->asic_prop.cfg_base_address + block_base + - FIELD_GET(HL_GLBL_ERR_ADDRESS_MASK, addr_val)); + "%s, addr %#llx\n", + hl_glbl_error_cause[i], + prop->cfg_base_address + block_base + + FIELD_GET(HL_GLBL_ERR_ADDRESS_MASK, addr_val)); } WREG32(glbl_err_cause, cause_val); diff --git a/drivers/accel/habanalabs/common/security.h b/drivers/accel/habanalabs/common/security.h index d7a3b3e82ea4..476f70687c09 100644 --- a/drivers/accel/habanalabs/common/security.h +++ b/drivers/accel/habanalabs/common/security.h @@ -13,8 +13,7 @@ struct hl_device; /* special blocks */ -#define HL_MAX_NUM_OF_GLBL_ERR_CAUSE 10 -#define HL_GLBL_ERR_ADDRESS_MASK GENMASK(11, 0) +#define HL_GLBL_ERR_ADDRESS_MASK GENMASK(11, 0) /* GLBL_ERR_ADDR register offset from the start of the block */ #define HL_GLBL_ERR_ADDR_OFFSET 0xF44 /* GLBL_ERR_CAUSE register offset from the start of the block */ diff --git a/drivers/accel/habanalabs/common/sysfs.c b/drivers/accel/habanalabs/common/sysfs.c index 8a9f98832157..9d58efa2ff38 100644 --- a/drivers/accel/habanalabs/common/sysfs.c +++ b/drivers/accel/habanalabs/common/sysfs.c @@ -142,8 +142,9 @@ static ssize_t cpld_ver_show(struct device *dev, struct device_attribute *attr, { struct hl_device *hdev = dev_get_drvdata(dev); - return sprintf(buf, "0x%08x\n", - le32_to_cpu(hdev->asic_prop.cpucp_info.cpld_version)); + return sprintf(buf, "0x%08x%08x\n", + le32_to_cpu(hdev->asic_prop.cpucp_info.cpld_timestamp), + le32_to_cpu(hdev->asic_prop.cpucp_info.cpld_version)); } static ssize_t cpucp_kernel_ver_show(struct device *dev, @@ -270,6 +271,9 @@ static ssize_t device_type_show(struct device *dev, case ASIC_GAUDI2C: str = "GAUDI2C"; break; + case ASIC_GAUDI2D: + str = "GAUDI2D"; + break; default: dev_err(hdev->dev, "Unrecognized ASIC type %d\n", hdev->asic_type); @@ -364,7 +368,7 @@ out: } static ssize_t eeprom_read_handler(struct file *filp, struct kobject *kobj, - struct bin_attribute *attr, char *buf, loff_t offset, + const struct bin_attribute *attr, char *buf, loff_t offset, size_t max_size) { struct device *dev = kobj_to_dev(kobj); @@ -439,10 +443,10 @@ static DEVICE_ATTR_RO(security_enabled); static DEVICE_ATTR_RO(module_id); static DEVICE_ATTR_RO(parent_device); -static struct bin_attribute bin_attr_eeprom = { +static const struct bin_attribute bin_attr_eeprom = { .attr = {.name = "eeprom", .mode = (0444)}, .size = PAGE_SIZE, - .read = eeprom_read_handler + .read_new = eeprom_read_handler }; static struct attribute *hl_dev_attrs[] = { @@ -468,14 +472,14 @@ static struct attribute *hl_dev_attrs[] = { NULL, }; -static struct bin_attribute *hl_dev_bin_attrs[] = { +static const struct bin_attribute *const hl_dev_bin_attrs[] = { &bin_attr_eeprom, NULL }; static struct attribute_group hl_dev_attr_group = { .attrs = hl_dev_attrs, - .bin_attrs = hl_dev_bin_attrs, + .bin_attrs_new = hl_dev_bin_attrs, }; static struct attribute_group hl_dev_clks_attr_group; diff --git a/drivers/accel/habanalabs/gaudi/gaudi.c b/drivers/accel/habanalabs/gaudi/gaudi.c index 53292d4c15c8..fa893a9b826e 100644 --- a/drivers/accel/habanalabs/gaudi/gaudi.c +++ b/drivers/accel/habanalabs/gaudi/gaudi.c @@ -614,8 +614,6 @@ static int gaudi_set_fixed_properties(struct hl_device *hdev) else prop->mmu_pgt_size = MMU_PAGE_TABLES_SIZE; prop->mmu_pte_size = HL_PTE_SIZE; - prop->mmu_hop_table_size = HOP_TABLE_SIZE_512_PTE; - prop->mmu_hop0_tables_total_size = HOP0_512_PTE_TABLES_TOTAL_SIZE; prop->dram_page_size = PAGE_SIZE_2MB; prop->device_mem_alloc_default_page_size = prop->dram_page_size; prop->dram_supports_virtual_memory = false; @@ -637,8 +635,8 @@ static int gaudi_set_fixed_properties(struct hl_device *hdev) prop->pmmu.num_hops = MMU_ARCH_5_HOPS; prop->pmmu.last_mask = LAST_MASK; /* TODO: will be duplicated until implementing per-MMU props */ - prop->pmmu.hop_table_size = prop->mmu_hop_table_size; - prop->pmmu.hop0_tables_total_size = prop->mmu_hop0_tables_total_size; + prop->pmmu.hop_table_size = HOP_TABLE_SIZE_512_PTE; + prop->pmmu.hop0_tables_total_size = HOP0_512_PTE_TABLES_TOTAL_SIZE; /* PMMU and HPMMU are the same except of page size */ memcpy(&prop->pmmu_huge, &prop->pmmu, sizeof(prop->pmmu)); @@ -649,6 +647,7 @@ static int gaudi_set_fixed_properties(struct hl_device *hdev) prop->dmmu.start_addr = (VA_HOST_SPACE_START + VA_HOST_SPACE_SIZE / 2); prop->dmmu.end_addr = VA_HOST_SPACE_END; prop->dmmu.page_size = PAGE_SIZE_2MB; + prop->dmmu.pgt_size = prop->mmu_pgt_size; prop->cfg_size = CFG_SIZE; prop->max_asid = MAX_ASID; @@ -1640,10 +1639,8 @@ static int gaudi_late_init(struct hl_device *hdev) } rc = hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_ENABLE_PCI_ACCESS, 0x0); - if (rc) { - dev_err(hdev->dev, "Failed to enable PCI access from CPU\n"); + if (rc) return rc; - } /* Scrub both SRAM and DRAM */ rc = hdev->asic_funcs->scrub_device_mem(hdev); @@ -3652,7 +3649,7 @@ static int gaudi_mmu_init(struct hl_device *hdev) for (i = 0 ; i < prop->max_asid ; i++) { hop0_addr = prop->mmu_pgt_addr + - (i * prop->mmu_hop_table_size); + (i * prop->dmmu.hop_table_size); rc = gaudi_mmu_update_asid_hop0_addr(hdev, i, hop0_addr); if (rc) { @@ -4155,13 +4152,7 @@ skip_reset: static int gaudi_suspend(struct hl_device *hdev) { - int rc; - - rc = hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0); - if (rc) - dev_err(hdev->dev, "Failed to disable PCI access from CPU\n"); - - return rc; + return hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0); } static int gaudi_resume(struct hl_device *hdev) diff --git a/drivers/accel/habanalabs/gaudi2/gaudi2.c b/drivers/accel/habanalabs/gaudi2/gaudi2.c index e0e5615ef9b0..a38b88baadf2 100644 --- a/drivers/accel/habanalabs/gaudi2/gaudi2.c +++ b/drivers/accel/habanalabs/gaudi2/gaudi2.c @@ -158,11 +158,13 @@ #define RAZWI_INITIATOR_ID_X_Y(xl, yl, xh) \ (RAZWI_INITIATOR_ID_X_Y_LOW(xl, yl) | RAZWI_INITIATOR_ID_X_HIGH(xh)) -#define PSOC_RAZWI_ENG_STR_SIZE 128 -#define PSOC_RAZWI_MAX_ENG_PER_RTR 5 +#define PSOC_RAZWI_ENG_STR_SIZE 128 +#define PSOC_RAZWI_MAX_ENG_PER_RTR 5 /* HW scrambles only bits 0-25 */ -#define HW_UNSCRAMBLED_BITS_MASK GENMASK_ULL(63, 26) +#define HW_UNSCRAMBLED_BITS_MASK GENMASK_ULL(63, 26) + +#define GAUDI2_GLBL_ERR_MAX_CAUSE_NUM 17 struct gaudi2_razwi_info { u32 axuser_xy; @@ -2308,11 +2310,26 @@ static int set_number_of_functional_hbms(struct hl_device *hdev) return 0; } +static bool gaudi2_is_edma_queue_id(u32 queue_id) +{ + + switch (queue_id) { + case GAUDI2_QUEUE_ID_DCORE0_EDMA_0_0...GAUDI2_QUEUE_ID_DCORE0_EDMA_1_3: + case GAUDI2_QUEUE_ID_DCORE1_EDMA_0_0...GAUDI2_QUEUE_ID_DCORE1_EDMA_1_3: + case GAUDI2_QUEUE_ID_DCORE2_EDMA_0_0...GAUDI2_QUEUE_ID_DCORE2_EDMA_1_3: + case GAUDI2_QUEUE_ID_DCORE3_EDMA_0_0...GAUDI2_QUEUE_ID_DCORE3_EDMA_1_3: + return true; + default: + return false; + } +} + static int gaudi2_set_dram_properties(struct hl_device *hdev) { struct asic_fixed_properties *prop = &hdev->asic_prop; - u32 basic_hbm_page_size; - int rc; + u64 hbm_drv_base_offset = 0, edma_pq_base_addr; + u32 basic_hbm_page_size, edma_idx = 0; + int rc, i; rc = set_number_of_functional_hbms(hdev); if (rc) @@ -2356,9 +2373,35 @@ static int gaudi2_set_dram_properties(struct hl_device *hdev) prop->dmmu.start_addr = prop->dram_base_address + (prop->dram_page_size * DIV_ROUND_UP_SECTOR_T(prop->dram_size, prop->dram_page_size)); - prop->dmmu.end_addr = prop->dmmu.start_addr + prop->dram_page_size * div_u64((VA_HBM_SPACE_END - prop->dmmu.start_addr), prop->dmmu.page_size); + /* + * Driver can't share an (48MB) HBM page with the F/W in order to prevent FW to block + * the driver part by range register, so it must start at the next (48MB) page + */ + hbm_drv_base_offset = roundup(CPU_FW_IMAGE_SIZE, prop->num_functional_hbms * SZ_8M); + + /* + * The NIC driver section size and the HMMU page tables section in the HBM needs + * to be the remaining size in the first dram page after taking into + * account the F/W image size + */ + + /* Reserve region in HBM for HMMU page tables */ + prop->mmu_pgt_addr = DRAM_PHYS_BASE + hbm_drv_base_offset + + ((prop->dram_page_size - hbm_drv_base_offset) - + (HMMU_PAGE_TABLES_SIZE + EDMA_PQS_SIZE + EDMA_SCRATCHPAD_SIZE)); + + /* Set EDMA PQs HBM addresses */ + edma_pq_base_addr = prop->mmu_pgt_addr + HMMU_PAGE_TABLES_SIZE; + + for (i = 0 ; i < GAUDI2_QUEUE_ID_CPU_PQ ; i++) { + if (gaudi2_is_edma_queue_id(i)) { + prop->hw_queues_props[i].q_dram_bd_address = edma_pq_base_addr + + (edma_idx * HL_QUEUE_SIZE_IN_BYTES); + edma_idx++; + } + } return 0; } @@ -2368,7 +2411,7 @@ static int gaudi2_set_fixed_properties(struct hl_device *hdev) struct asic_fixed_properties *prop = &hdev->asic_prop; struct hw_queue_properties *q_props; u32 num_sync_stream_queues = 0; - int i; + int i, rc; prop->max_queues = GAUDI2_QUEUE_ID_SIZE; prop->hw_queues_props = kcalloc(prop->max_queues, sizeof(struct hw_queue_properties), @@ -2391,6 +2434,9 @@ static int gaudi2_set_fixed_properties(struct hl_device *hdev) } q_props[i].cb_alloc_flags = CB_ALLOC_USER; + + if (gaudi2_is_edma_queue_id(i)) + q_props[i].dram_bd = 1; } q_props[GAUDI2_QUEUE_ID_CPU_PQ].type = QUEUE_TYPE_CPU; @@ -2419,46 +2465,43 @@ static int gaudi2_set_fixed_properties(struct hl_device *hdev) prop->rotator_enabled_mask = BIT(NUM_OF_ROT) - 1; - if (hdev->pldm) - prop->mmu_pgt_size = 0x800000; /* 8MB */ - else - prop->mmu_pgt_size = MMU_PAGE_TABLES_INITIAL_SIZE; + prop->max_asid = 2; + prop->dmmu.pgt_size = HMMU_PAGE_TABLES_SIZE; prop->mmu_pte_size = HL_PTE_SIZE; - prop->mmu_hop_table_size = HOP_TABLE_SIZE_512_PTE; - prop->mmu_hop0_tables_total_size = HOP0_512_PTE_TABLES_TOTAL_SIZE; prop->dmmu.hop_shifts[MMU_HOP0] = DHOP0_SHIFT; prop->dmmu.hop_shifts[MMU_HOP1] = DHOP1_SHIFT; prop->dmmu.hop_shifts[MMU_HOP2] = DHOP2_SHIFT; prop->dmmu.hop_shifts[MMU_HOP3] = DHOP3_SHIFT; - prop->dmmu.hop_shifts[MMU_HOP4] = DHOP4_SHIFT; prop->dmmu.hop_masks[MMU_HOP0] = DHOP0_MASK; prop->dmmu.hop_masks[MMU_HOP1] = DHOP1_MASK; prop->dmmu.hop_masks[MMU_HOP2] = DHOP2_MASK; prop->dmmu.hop_masks[MMU_HOP3] = DHOP3_MASK; - prop->dmmu.hop_masks[MMU_HOP4] = DHOP4_MASK; prop->dmmu.page_size = PAGE_SIZE_1GB; - prop->dmmu.num_hops = MMU_ARCH_6_HOPS; + prop->dmmu.num_hops = MMU_ARCH_4_HOPS; prop->dmmu.last_mask = LAST_MASK; - prop->dmmu.host_resident = 1; - prop->dmmu.hop_table_size = prop->mmu_hop_table_size; - prop->dmmu.hop0_tables_total_size = prop->mmu_hop0_tables_total_size; + prop->dmmu.host_resident = 0; + prop->dmmu.hop_table_size = HOP_TABLE_SIZE_512_PTE; + prop->dmmu.hop0_tables_total_size = HOP_TABLE_SIZE_512_PTE * prop->max_asid; - /* - * this is done in order to be able to validate FW descriptor (i.e. validating that - * the addresses and allocated space for FW image does not cross memory bounds). - * for this reason we set the DRAM size to the minimum possible and later it will - * be modified according to what reported in the cpucp info packet + /* As we need to set the pgt address in dram for HMMU init so we cannot + * wait to the fw cpucp info to set the dram props as mmu init comes before + * hw init */ - prop->dram_size = (GAUDI2_HBM_NUM - 1) * SZ_16G; + rc = hdev->asic_funcs->set_dram_properties(hdev); + if (rc) + goto free_qprops; + + prop->mmu_pgt_size = PMMU_PAGE_TABLES_SIZE; + prop->pmmu.pgt_size = prop->mmu_pgt_size; hdev->pmmu_huge_range = true; prop->pmmu.host_resident = 1; prop->pmmu.num_hops = MMU_ARCH_6_HOPS; prop->pmmu.last_mask = LAST_MASK; - prop->pmmu.hop_table_size = prop->mmu_hop_table_size; - prop->pmmu.hop0_tables_total_size = prop->mmu_hop0_tables_total_size; + prop->pmmu.hop_table_size = HOP_TABLE_SIZE_512_PTE; + prop->pmmu.hop0_tables_total_size = HOP_TABLE_SIZE_512_PTE * prop->max_asid; prop->hints_host_reserved_va_range.start_addr = RESERVED_VA_FOR_VIRTUAL_MSIX_DOORBELL_START; prop->hints_host_reserved_va_range.end_addr = RESERVED_VA_RANGE_FOR_ARC_ON_HOST_END; @@ -2516,7 +2559,6 @@ static int gaudi2_set_fixed_properties(struct hl_device *hdev) prop->max_num_of_engines = GAUDI2_ENGINE_ID_SIZE; prop->num_engine_cores = CPU_ID_MAX; prop->cfg_size = CFG_SIZE; - prop->max_asid = MAX_ASID; prop->num_of_events = GAUDI2_EVENT_SIZE; prop->supports_engine_modes = true; @@ -2559,7 +2601,13 @@ static int gaudi2_set_fixed_properties(struct hl_device *hdev) prop->hbw_flush_reg = mmPCIE_WRAP_SPECIAL_GLBL_SPARE_0; + prop->supports_advanced_cpucp_rc = true; + return 0; + +free_qprops: + kfree(prop->hw_queues_props); + return rc; } static int gaudi2_pci_bars_map(struct hl_device *hdev) @@ -3033,6 +3081,25 @@ static int gaudi2_fetch_psoc_frequency(struct hl_device *hdev) return 0; } +static int gaudi2_mmu_clear_pgt_range(struct hl_device *hdev) +{ + struct gaudi2_device *gaudi2 = hdev->asic_specific; + struct asic_fixed_properties *prop = &hdev->asic_prop; + int rc; + + if (!(gaudi2->hw_cap_initialized & HW_CAP_MMU_MASK)) + return 0; + + if (prop->dmmu.host_resident) + return 0; + + rc = gaudi2_memset_device_memory(hdev, prop->mmu_pgt_addr, prop->dmmu.pgt_size, 0); + if (rc) + dev_err(hdev->dev, "Failed to clear mmu pgt"); + + return rc; +} + static int gaudi2_early_init(struct hl_device *hdev) { struct asic_fixed_properties *prop = &hdev->asic_prop; @@ -3243,14 +3310,10 @@ static int gaudi2_late_init(struct hl_device *hdev) struct gaudi2_device *gaudi2 = hdev->asic_specific; int rc; - hdev->asic_prop.supports_advanced_cpucp_rc = true; - rc = hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_ENABLE_PCI_ACCESS, gaudi2->virt_msix_db_dma_addr); - if (rc) { - dev_err(hdev->dev, "Failed to enable PCI access from CPU\n"); + if (rc) return rc; - } rc = gaudi2_fetch_psoc_frequency(hdev); if (rc) { @@ -3258,6 +3321,12 @@ static int gaudi2_late_init(struct hl_device *hdev) goto disable_pci_access; } + rc = gaudi2_mmu_clear_pgt_range(hdev); + if (rc) { + dev_err(hdev->dev, "Failed to clear MMU page tables range\n"); + goto disable_pci_access; + } + gaudi2_init_arcs(hdev); rc = gaudi2_scrub_arcs_dccm(hdev); @@ -3518,7 +3587,7 @@ static int gaudi2_special_blocks_config(struct hl_device *hdev) int i, rc; /* Configure Special blocks */ - prop->glbl_err_cause_num = GAUDI2_NUM_OF_GLBL_ERR_CAUSE; + prop->glbl_err_max_cause_num = GAUDI2_GLBL_ERR_MAX_CAUSE_NUM; prop->num_of_special_blocks = ARRAY_SIZE(gaudi2_special_blocks); prop->special_blocks = kmalloc_array(prop->num_of_special_blocks, sizeof(*prop->special_blocks), GFP_KERNEL); @@ -3697,13 +3766,7 @@ static int gaudi2_sw_init(struct hl_device *hdev) spin_lock_init(&gaudi2->hw_queues_lock); - gaudi2->scratchpad_kernel_address = hl_asic_dma_alloc_coherent(hdev, PAGE_SIZE, - &gaudi2->scratchpad_bus_address, - GFP_KERNEL | __GFP_ZERO); - if (!gaudi2->scratchpad_kernel_address) { - rc = -ENOMEM; - goto free_virt_msix_db_mem; - } + gaudi2->scratchpad_bus_address = prop->mmu_pgt_addr + HMMU_PAGE_TABLES_SIZE + EDMA_PQS_SIZE; gaudi2_user_mapped_blocks_init(hdev); @@ -3718,7 +3781,7 @@ static int gaudi2_sw_init(struct hl_device *hdev) prop->supports_compute_reset = true; /* Event queue sanity check added in FW version 1.11 */ - if (hl_is_fw_sw_ver_below(hdev, 1, 11)) + if (hl_fw_version_cmp(hdev, 1, 11, 0) < 0) hdev->event_queue.check_eqe_index = false; else hdev->event_queue.check_eqe_index = true; @@ -3727,19 +3790,18 @@ static int gaudi2_sw_init(struct hl_device *hdev) rc = gaudi2_special_blocks_iterator_config(hdev); if (rc) - goto free_scratchpad_mem; + goto free_virt_msix_db_mem; rc = gaudi2_test_queues_msgs_alloc(hdev); if (rc) goto special_blocks_free; + hdev->heartbeat_debug_info.cpu_queue_id = GAUDI2_QUEUE_ID_CPU_PQ; + return 0; special_blocks_free: gaudi2_special_blocks_iterator_free(hdev); -free_scratchpad_mem: - hl_asic_dma_free_coherent(hdev, PAGE_SIZE, gaudi2->scratchpad_kernel_address, - gaudi2->scratchpad_bus_address); free_virt_msix_db_mem: hl_cpu_accessible_dma_pool_free(hdev, prop->pmmu.page_size, gaudi2->virt_msix_db_cpu_addr); free_cpu_accessible_dma_pool: @@ -3770,9 +3832,6 @@ static int gaudi2_sw_fini(struct hl_device *hdev) hl_asic_dma_free_coherent(hdev, HL_CPU_ACCESSIBLE_MEM_SIZE, hdev->cpu_accessible_dma_mem, hdev->cpu_accessible_dma_address); - hl_asic_dma_free_coherent(hdev, PAGE_SIZE, gaudi2->scratchpad_kernel_address, - gaudi2->scratchpad_bus_address); - dma_pool_destroy(hdev->dma_pool); kfree(gaudi2); @@ -4254,6 +4313,8 @@ static int gaudi2_enable_msix(struct hl_device *hdev) if (gaudi2->hw_cap_initialized & HW_CAP_MSIX) return 0; + hl_init_cpu_for_irq(hdev); + rc = pci_alloc_irq_vectors(hdev->pdev, GAUDI2_MSIX_ENTRIES, GAUDI2_MSIX_ENTRIES, PCI_IRQ_MSIX); if (rc < 0) { @@ -4307,6 +4368,7 @@ static int gaudi2_enable_msix(struct hl_device *hdev) i++, j++, user_irq_init_cnt++) { irq = pci_irq_vector(hdev->pdev, i); + hl_set_irq_affinity(hdev, irq); rc = request_irq(irq, hl_irq_user_interrupt_handler, 0, gaudi2_irq_name(i), &hdev->user_interrupt[j]); if (rc) { @@ -4333,6 +4395,7 @@ free_user_irq: i < GAUDI2_IRQ_NUM_USER_FIRST + user_irq_init_cnt ; i++, j++) { irq = pci_irq_vector(hdev->pdev, i); + irq_set_affinity_and_hint(irq, NULL); free_irq(irq, &hdev->user_interrupt[j]); } irq = pci_irq_vector(hdev->pdev, GAUDI2_IRQ_NUM_UNEXPECTED_ERROR); @@ -4413,6 +4476,7 @@ static void gaudi2_disable_msix(struct hl_device *hdev) k < hdev->asic_prop.user_interrupt_count ; i++, j++, k++) { irq = pci_irq_vector(hdev->pdev, i); + irq_set_affinity_and_hint(irq, NULL); free_irq(irq, &hdev->user_interrupt[j]); } @@ -4957,10 +5021,17 @@ static void gaudi2_init_qman_pq(struct hl_device *hdev, u32 reg_base, q = &hdev->kernel_queues[queue_id_base + pq_id]; pq_offset = pq_id * 4; - WREG32(reg_base + QM_PQ_BASE_LO_0_OFFSET + pq_offset, - lower_32_bits(q->bus_address)); - WREG32(reg_base + QM_PQ_BASE_HI_0_OFFSET + pq_offset, - upper_32_bits(q->bus_address)); + if (q->dram_bd) { + WREG32(reg_base + QM_PQ_BASE_LO_0_OFFSET + pq_offset, + lower_32_bits(q->pq_dram_address)); + WREG32(reg_base + QM_PQ_BASE_HI_0_OFFSET + pq_offset, + upper_32_bits(q->pq_dram_address)); + } else { + WREG32(reg_base + QM_PQ_BASE_LO_0_OFFSET + pq_offset, + lower_32_bits(q->bus_address)); + WREG32(reg_base + QM_PQ_BASE_HI_0_OFFSET + pq_offset, + upper_32_bits(q->bus_address)); + } WREG32(reg_base + QM_PQ_SIZE_0_OFFSET + pq_offset, ilog2(HL_QUEUE_LENGTH)); WREG32(reg_base + QM_PQ_PI_0_OFFSET + pq_offset, 0); WREG32(reg_base + QM_PQ_CI_0_OFFSET + pq_offset, 0); @@ -5847,7 +5918,8 @@ static int gaudi2_mmu_invalidate_cache_range(struct hl_device *hdev, bool is_har return rc; } -static int gaudi2_mmu_update_hop0_addr(struct hl_device *hdev, u32 stlb_base) +static int gaudi2_mmu_update_hop0_addr(struct hl_device *hdev, u32 stlb_base, + bool host_resident_pgt) { struct asic_fixed_properties *prop = &hdev->asic_prop; u64 hop0_addr; @@ -5859,7 +5931,11 @@ static int gaudi2_mmu_update_hop0_addr(struct hl_device *hdev, u32 stlb_base) max_asid = min((u32) 8, max_asid); for (asid = 0 ; asid < max_asid ; asid++) { - hop0_addr = hdev->mmu_priv.hr.mmu_asid_hop0[asid].phys_addr; + if (host_resident_pgt) + hop0_addr = hdev->mmu_priv.hr.mmu_asid_hop0[asid].phys_addr; + else + hop0_addr = prop->mmu_pgt_addr + (asid * prop->dmmu.hop_table_size); + rc = gaudi2_mmu_update_asid_hop0_addr(hdev, stlb_base, asid, hop0_addr); if (rc) { dev_err(hdev->dev, "failed to set hop0 addr for asid %d\n", asid); @@ -5870,7 +5946,8 @@ static int gaudi2_mmu_update_hop0_addr(struct hl_device *hdev, u32 stlb_base) return 0; } -static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base, u32 stlb_base) +static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base, u32 stlb_base, + bool host_resident_pgt) { u32 status, timeout_usec; int rc; @@ -5893,7 +5970,7 @@ static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base, u32 stlb if (rc) dev_notice_ratelimited(hdev->dev, "Timeout when waiting for MMU SRAM init\n"); - rc = gaudi2_mmu_update_hop0_addr(hdev, stlb_base); + rc = gaudi2_mmu_update_hop0_addr(hdev, stlb_base, host_resident_pgt); if (rc) return rc; @@ -5917,6 +5994,7 @@ static int gaudi2_mmu_init_common(struct hl_device *hdev, u32 mmu_base, u32 stlb static int gaudi2_pci_mmu_init(struct hl_device *hdev) { + struct asic_fixed_properties *prop = &hdev->asic_prop; struct gaudi2_device *gaudi2 = hdev->asic_specific; u32 mmu_base, stlb_base; int rc; @@ -5956,7 +6034,7 @@ static int gaudi2_pci_mmu_init(struct hl_device *hdev) WREG32(mmu_base + MMU_SPI_SEI_MASK_OFFSET, GAUDI2_PMMU_SPI_SEI_ENABLE_MASK); - rc = gaudi2_mmu_init_common(hdev, mmu_base, stlb_base); + rc = gaudi2_mmu_init_common(hdev, mmu_base, stlb_base, prop->pmmu.host_resident); if (rc) return rc; @@ -6008,7 +6086,7 @@ static int gaudi2_dcore_hmmu_init(struct hl_device *hdev, int dcore_id, WREG32(mmu_base + MMU_SPI_SEI_MASK_OFFSET, GAUDI2_HMMU_SPI_SEI_ENABLE_MASK); - rc = gaudi2_mmu_init_common(hdev, mmu_base, stlb_base); + rc = gaudi2_mmu_init_common(hdev, mmu_base, stlb_base, prop->dmmu.host_resident); if (rc) return rc; @@ -6236,26 +6314,6 @@ static void gaudi2_execute_hard_reset(struct hl_device *hdev) WREG32(mmPSOC_RESET_CONF_SW_ALL_RST, 1); } -static int gaudi2_get_soft_rst_done_indication(struct hl_device *hdev, u32 poll_timeout_us) -{ - int i, rc = 0; - u32 reg_val; - - for (i = 0 ; i < GAUDI2_RESET_POLL_CNT ; i++) - rc = hl_poll_timeout( - hdev, - mmCPU_RST_STATUS_TO_HOST, - reg_val, - reg_val == CPU_RST_STATUS_SOFT_RST_DONE, - 1000, - poll_timeout_us); - - if (rc) - dev_err(hdev->dev, "Timeout while waiting for FW to complete soft reset (0x%x)\n", - reg_val); - return rc; -} - /** * gaudi2_execute_soft_reset - execute soft reset by driver/FW * @@ -6268,23 +6326,8 @@ static int gaudi2_get_soft_rst_done_indication(struct hl_device *hdev, u32 poll_ static int gaudi2_execute_soft_reset(struct hl_device *hdev, bool driver_performs_reset, u32 poll_timeout_us) { - int rc; - - if (!driver_performs_reset) { - if (hl_is_fw_sw_ver_below(hdev, 1, 10)) { - /* set SP to indicate reset request sent to FW */ - WREG32(mmCPU_RST_STATUS_TO_HOST, CPU_RST_STATUS_NA); - - WREG32(mmGIC_HOST_SOFT_RST_IRQ_POLL_REG, - gaudi2_irq_map_table[GAUDI2_EVENT_CPU_SOFT_RESET].cpu_id); - - /* wait for f/w response */ - rc = gaudi2_get_soft_rst_done_indication(hdev, poll_timeout_us); - } else { - rc = hl_fw_send_soft_reset(hdev); - } - return rc; - } + if (!driver_performs_reset) + return hl_fw_send_soft_reset(hdev); /* Block access to engines, QMANs and SM during reset, these * RRs will be reconfigured after soft reset. @@ -6424,13 +6467,7 @@ skip_reset: static int gaudi2_suspend(struct hl_device *hdev) { - int rc; - - rc = hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0); - if (rc) - dev_err(hdev->dev, "Failed to disable PCI access from CPU\n"); - - return rc; + return hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0); } static int gaudi2_resume(struct hl_device *hdev) @@ -7046,7 +7083,7 @@ static int gaudi2_test_queues(struct hl_device *hdev) /* send test message on all enabled Qs */ for (i = GAUDI2_QUEUE_ID_PDMA_0_0 ; i < GAUDI2_QUEUE_ID_CPU_PQ; i++) { - if (!gaudi2_is_queue_enabled(hdev, i)) + if (!gaudi2_is_queue_enabled(hdev, i) || gaudi2_is_edma_queue_id(i)) continue; msg_info = &gaudi2->queues_test_info[i - GAUDI2_QUEUE_ID_PDMA_0_0]; @@ -7063,7 +7100,7 @@ static int gaudi2_test_queues(struct hl_device *hdev) /* verify that all messages were processed */ for (i = GAUDI2_QUEUE_ID_PDMA_0_0 ; i < GAUDI2_QUEUE_ID_CPU_PQ; i++) { - if (!gaudi2_is_queue_enabled(hdev, i)) + if (!gaudi2_is_queue_enabled(hdev, i) || gaudi2_is_edma_queue_id(i)) continue; rc = gaudi2_test_queue_wait_completion(hdev, i, sob_val); @@ -7836,7 +7873,7 @@ static bool gaudi2_handle_ecc_event(struct hl_device *hdev, u16 event_type, bool has_block_id = false; u16 block_id; - if (!hl_is_fw_sw_ver_below(hdev, 1, 12)) + if (hl_fw_version_cmp(hdev, 1, 12, 0) >= 0) has_block_id = true; ecc_address = le64_to_cpu(ecc_data->ecc_address); @@ -8087,13 +8124,7 @@ static void gaudi2_ack_module_razwi_event_handler(struct hl_device *hdev, } hbw_rtr_id = gaudi2_tpc_initiator_hbw_rtr_id[module_idx]; - - if (hl_is_fw_sw_ver_below(hdev, 1, 9) && - !hdev->asic_prop.fw_security_enabled && - ((module_idx == 0) || (module_idx == 1))) - lbw_rtr_id = DCORE0_RTR0; - else - lbw_rtr_id = gaudi2_tpc_initiator_lbw_rtr_id[module_idx]; + lbw_rtr_id = gaudi2_tpc_initiator_lbw_rtr_id[module_idx]; break; case RAZWI_MME: sprintf(initiator_name, "MME_%u", module_idx); @@ -8907,9 +8938,6 @@ static int gaudi2_print_pcie_addr_dec_info(struct hl_device *hdev, u16 event_typ u32 error_count = 0; int i; - gaudi2_print_event(hdev, event_type, true, - "intr_cause_data: %#llx", intr_cause_data); - for (i = 0 ; i < GAUDI2_NUM_OF_PCIE_ADDR_DEC_ERR_CAUSE ; i++) { if (!(intr_cause_data & BIT_ULL(i))) continue; @@ -8918,15 +8946,16 @@ static int gaudi2_print_pcie_addr_dec_info(struct hl_device *hdev, u16 event_typ "err cause: %s", gaudi2_pcie_addr_dec_error_cause[i]); error_count++; - /* - * Always check for LBW and HBW additional info as the indication itself is - * sometimes missing - */ + switch (intr_cause_data & BIT_ULL(i)) { + case PCIE_WRAP_PCIE_IC_SEI_INTR_IND_AXI_LBW_ERR_INTR_MASK: + hl_check_for_glbl_errors(hdev); + break; + case PCIE_WRAP_PCIE_IC_SEI_INTR_IND_BAD_ACCESS_INTR_MASK: + gaudi2_print_pcie_mstr_rr_mstr_if_razwi_info(hdev, event_mask); + break; + } } - hl_check_for_glbl_errors(hdev); - gaudi2_print_pcie_mstr_rr_mstr_if_razwi_info(hdev, event_mask); - return error_count; } @@ -8983,7 +9012,6 @@ static void gaudi2_handle_page_error(struct hl_device *hdev, u64 mmu_base, bool if (is_pmmu) { dev_err_ratelimited(hdev->dev, "PMMU page fault on va 0x%llx\n", addr); } else { - addr = gaudi2_mmu_descramble_addr(hdev, addr); addr &= HW_UNSCRAMBLED_BITS_MASK; dev_err_ratelimited(hdev->dev, "HMMU page fault on va range 0x%llx - 0x%llx\n", @@ -9235,8 +9263,8 @@ static int gaudi2_handle_mmu_spi_sei_err(struct hl_device *hdev, u16 event_type, static bool gaudi2_hbm_sei_handle_read_err(struct hl_device *hdev, struct hl_eq_hbm_sei_read_err_intr_info *rd_err_data, u32 err_cnt) { + bool require_hard_reset = false; u32 addr, beat, beat_shift; - bool rc = false; dev_err_ratelimited(hdev->dev, "READ ERROR count: ECC SERR: %d, ECC DERR: %d, RD_PARITY: %d\n", @@ -9268,7 +9296,7 @@ static bool gaudi2_hbm_sei_handle_read_err(struct hl_device *hdev, beat, le32_to_cpu(rd_err_data->dbg_rd_err_dm), le32_to_cpu(rd_err_data->dbg_rd_err_syndrome)); - rc |= true; + require_hard_reset = true; } beat_shift = beat * HBM_RD_ERR_BEAT_SHIFT; @@ -9281,7 +9309,7 @@ static bool gaudi2_hbm_sei_handle_read_err(struct hl_device *hdev, (le32_to_cpu(rd_err_data->dbg_rd_err_misc) & (HBM_RD_ERR_PAR_DATA_BEAT0_MASK << beat_shift)) >> (HBM_RD_ERR_PAR_DATA_BEAT0_SHIFT + beat_shift)); - rc |= true; + require_hard_reset = true; } dev_err_ratelimited(hdev->dev, "Beat%d DQ data:\n", beat); @@ -9291,7 +9319,7 @@ static bool gaudi2_hbm_sei_handle_read_err(struct hl_device *hdev, le32_to_cpu(rd_err_data->dbg_rd_err_data[beat * 2 + 1])); } - return rc; + return require_hard_reset; } static void gaudi2_hbm_sei_print_wr_par_info(struct hl_device *hdev, @@ -9514,25 +9542,17 @@ static int gaudi2_handle_pcie_p2p_msix(struct hl_device *hdev, u16 event_type) static int gaudi2_handle_pcie_drain(struct hl_device *hdev, struct hl_eq_pcie_drain_ind_data *drain_data) { - u64 lbw_rd, lbw_wr, hbw_rd, hbw_wr, cause, error_count = 0; + u64 cause, error_count = 0; cause = le64_to_cpu(drain_data->intr_cause.intr_cause_data); - lbw_rd = le64_to_cpu(drain_data->drain_rd_addr_lbw); - lbw_wr = le64_to_cpu(drain_data->drain_wr_addr_lbw); - hbw_rd = le64_to_cpu(drain_data->drain_rd_addr_hbw); - hbw_wr = le64_to_cpu(drain_data->drain_wr_addr_hbw); if (cause & BIT_ULL(0)) { - dev_err_ratelimited(hdev->dev, - "PCIE AXI drain LBW completed, read_err %u, write_err %u\n", - !!lbw_rd, !!lbw_wr); + dev_err_ratelimited(hdev->dev, "PCIE AXI drain LBW completed\n"); error_count++; } if (cause & BIT_ULL(1)) { - dev_err_ratelimited(hdev->dev, - "PCIE AXI drain HBW completed, raddr %#llx, waddr %#llx\n", - hbw_rd, hbw_wr); + dev_err_ratelimited(hdev->dev, "PCIE AXI drain HBW completed\n"); error_count++; } @@ -9757,11 +9777,6 @@ static u16 event_id_to_engine_id(struct hl_device *hdev, u16 event_type) return U16_MAX; } -static void hl_eq_heartbeat_event_handle(struct hl_device *hdev) -{ - hdev->eq_heartbeat_received = true; -} - static void gaudi2_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_entry) { struct gaudi2_device *gaudi2 = hdev->asic_specific; @@ -9983,6 +9998,7 @@ static void gaudi2_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_ent if (gaudi2_handle_hbm_mc_sei_err(hdev, event_type, &eq_entry->sei_data)) { reset_flags |= HL_DRV_RESET_FW_FATAL_ERR; reset_required = true; + is_critical = eq_entry->sei_data.hdr.is_critical; } error_count++; break; @@ -10003,7 +10019,7 @@ static void gaudi2_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_ent error_count = gaudi2_handle_pcie_drain(hdev, &eq_entry->pcie_drain_ind_data); reset_flags |= HL_DRV_RESET_FW_FATAL_ERR; event_mask |= HL_NOTIFIER_EVENT_GENERAL_HW_ERR; - if (hl_is_fw_sw_ver_equal_or_greater(hdev, 1, 13)) + if (hl_fw_version_cmp(hdev, 1, 13, 0) >= 0) is_critical = true; break; @@ -10214,8 +10230,7 @@ static void gaudi2_handle_eqe(struct hl_device *hdev, struct hl_eq_entry *eq_ent gaudi2_print_event(hdev, event_type, true, "No error cause for H/W event %u", event_type); - if ((gaudi2_irq_map_table[event_type].reset != EVENT_RESET_TYPE_NONE) || - reset_required) { + if ((gaudi2_irq_map_table[event_type].reset != EVENT_RESET_TYPE_NONE) || reset_required) { if (reset_required || (gaudi2_irq_map_table[event_type].reset == EVENT_RESET_TYPE_HARD)) reset_flags |= HL_DRV_RESET_HARD; @@ -10250,11 +10265,11 @@ reset_device: } static int gaudi2_memset_memory_chunk_using_edma_qm(struct hl_device *hdev, - struct packet_lin_dma *lin_dma_pkt, dma_addr_t pkt_dma_addr, - u32 hw_queue_id, u32 size, u64 addr, u32 val) + struct packet_lin_dma *lin_dma_pkt, + u64 phys_addr, u32 hw_queue_id, u32 size, u64 addr, u32 val) { u32 ctl, pkt_size; - int rc = 0; + int rc = 0, i; ctl = FIELD_PREP(GAUDI2_PKT_CTL_OPCODE_MASK, PACKET_LIN_DMA); ctl |= FIELD_PREP(GAUDI2_PKT_LIN_DMA_CTL_MEMSET_MASK, 1); @@ -10268,9 +10283,20 @@ static int gaudi2_memset_memory_chunk_using_edma_qm(struct hl_device *hdev, pkt_size = sizeof(struct packet_lin_dma); - rc = hl_hw_queue_send_cb_no_cmpl(hdev, hw_queue_id, pkt_size, pkt_dma_addr); + for (i = 0; i < 3; i++) { + rc = hdev->asic_funcs->access_dev_mem(hdev, PCI_REGION_DRAM, + phys_addr + (i * sizeof(u64)), + ((u64 *)(lin_dma_pkt)) + i, DEBUGFS_WRITE64); + if (rc) { + dev_err(hdev->dev, "Failed to copy lin_dma packet to HBM (%#llx)\n", + phys_addr); + return rc; + } + } + + rc = hl_hw_queue_send_cb_no_cmpl(hdev, hw_queue_id, pkt_size, phys_addr); if (rc) - dev_err(hdev->dev, "Failed to send lin dma packet to H/W queue %d\n", + dev_err(hdev->dev, "Failed to send lin_dma packet to H/W queue %d\n", hw_queue_id); return rc; @@ -10283,12 +10309,11 @@ static int gaudi2_memset_device_memory(struct hl_device *hdev, u64 addr, u64 siz GAUDI2_QUEUE_ID_DCORE2_EDMA_0_0, GAUDI2_QUEUE_ID_DCORE3_EDMA_0_0}; u32 chunk_size, dcore, edma_idx, sob_offset, sob_addr, comp_val, - old_mmubp, mmubp, num_of_pkts, busy, pkt_size; + old_mmubp, mmubp, num_of_pkts, busy, pkt_size, cb_len; u64 comp_addr, cur_addr = addr, end_addr = addr + size; struct asic_fixed_properties *prop = &hdev->asic_prop; + int rc = 0, dma_num = 0, i; void *lin_dma_pkts_arr; - dma_addr_t pkt_dma_addr; - int rc = 0, dma_num = 0; if (prop->edma_enabled_mask == 0) { dev_info(hdev->dev, "non of the EDMA engines is enabled - skip dram scrubbing\n"); @@ -10306,9 +10331,19 @@ static int gaudi2_memset_device_memory(struct hl_device *hdev, u64 addr, u64 siz /* Calculate how many lin dma pkts we'll need */ num_of_pkts = div64_u64(round_up(size, SZ_2G), SZ_2G); pkt_size = sizeof(struct packet_lin_dma); + cb_len = pkt_size * num_of_pkts; - lin_dma_pkts_arr = hl_asic_dma_alloc_coherent(hdev, pkt_size * num_of_pkts, - &pkt_dma_addr, GFP_KERNEL); + /* + * if we're not scrubing HMMU or NIC reserved sections in hbm, + * then it the scrubing of the user section, as we use the start of the user section + * to store the CB of the EDMA QM, so shift the start address of the scrubbing accordingly + * and scrub the CB section before leaving this function. + */ + if ((addr >= prop->dram_user_base_address) && + (addr < prop->dram_user_base_address + cb_len)) + cur_addr += (prop->dram_user_base_address + cb_len) - addr; + + lin_dma_pkts_arr = kvcalloc(num_of_pkts, pkt_size, GFP_KERNEL); if (!lin_dma_pkts_arr) return -ENOMEM; @@ -10354,7 +10389,7 @@ static int gaudi2_memset_device_memory(struct hl_device *hdev, u64 addr, u64 siz rc = gaudi2_memset_memory_chunk_using_edma_qm(hdev, (struct packet_lin_dma *)lin_dma_pkts_arr + dma_num, - pkt_dma_addr + dma_num * pkt_size, + prop->dram_user_base_address + (dma_num * pkt_size), edma_queues_id[dcore] + edma_idx * 4, chunk_size, cur_addr, val); if (rc) @@ -10363,14 +10398,16 @@ static int gaudi2_memset_device_memory(struct hl_device *hdev, u64 addr, u64 siz dma_num++; cur_addr += chunk_size; if (cur_addr == end_addr) - break; + goto edma_wait; } } } +edma_wait: rc = hl_poll_timeout(hdev, sob_addr, busy, (busy == dma_num), 1000, 1000000); if (rc) { - dev_err(hdev->dev, "DMA Timeout during HBM scrubbing\n"); + dev_err(hdev->dev, "DMA Timeout during HBM scrubbing(sob: 0x%x, dma_num: 0x%x)\n", + busy, dma_num); goto end; } end: @@ -10391,8 +10428,16 @@ end: } } + memset(lin_dma_pkts_arr, 0, sizeof(u64)); + + /* Zero the HBM area where we copied the CB */ + for (i = 0; i < cb_len / sizeof(u64); i += sizeof(u64)) + rc = hdev->asic_funcs->access_dev_mem(hdev, PCI_REGION_DRAM, + prop->dram_user_base_address + i, + (u64 *)(lin_dma_pkts_arr), DEBUGFS_WRITE64); WREG32(sob_addr, 0); - hl_asic_dma_free_coherent(hdev, pkt_size * num_of_pkts, lin_dma_pkts_arr, pkt_dma_addr); + + kfree(lin_dma_pkts_arr); return rc; } @@ -11450,7 +11495,7 @@ static int gaudi2_mmu_get_real_page_size(struct hl_device *hdev, struct hl_mmu_p return 0; page_size_err: - dev_err(hdev->dev, "page size of %u is not %uKB aligned, can't map\n", + dev_err(hdev->dev, "page size of 0x%X is not 0x%X aligned, can't map\n", page_size, mmu_prop->page_size >> 10); return -EFAULT; } @@ -11470,6 +11515,29 @@ int gaudi2_send_device_activity(struct hl_device *hdev, bool open) return hl_fw_send_device_activity(hdev, open); } +static u64 gaudi2_read_pte(struct hl_device *hdev, u64 addr) +{ + struct gaudi2_device *gaudi2 = hdev->asic_specific; + u64 val; + + if (hdev->reset_info.hard_reset_pending) + return U64_MAX; + + val = readq(hdev->pcie_bar[DRAM_BAR_ID] + (addr - gaudi2->dram_bar_cur_addr)); + + return val; +} + +static void gaudi2_write_pte(struct hl_device *hdev, u64 addr, u64 val) +{ + struct gaudi2_device *gaudi2 = hdev->asic_specific; + + if (hdev->reset_info.hard_reset_pending) + return; + + writeq(val, hdev->pcie_bar[DRAM_BAR_ID] + (addr - gaudi2->dram_bar_cur_addr)); +} + static const struct hl_asic_funcs gaudi2_funcs = { .early_init = gaudi2_early_init, .early_fini = gaudi2_early_fini, @@ -11506,8 +11574,8 @@ static const struct hl_asic_funcs gaudi2_funcs = { .add_device_attr = gaudi2_add_device_attr, .handle_eqe = gaudi2_handle_eqe, .get_events_stat = gaudi2_get_events_stat, - .read_pte = NULL, - .write_pte = NULL, + .read_pte = gaudi2_read_pte, + .write_pte = gaudi2_write_pte, .mmu_invalidate_cache = gaudi2_mmu_invalidate_cache, .mmu_invalidate_cache_range = gaudi2_mmu_invalidate_cache_range, .mmu_prefetch_cache_range = NULL, diff --git a/drivers/accel/habanalabs/gaudi2/gaudi2P.h b/drivers/accel/habanalabs/gaudi2/gaudi2P.h index 9b9eef0d97d6..05117272cac7 100644 --- a/drivers/accel/habanalabs/gaudi2/gaudi2P.h +++ b/drivers/accel/habanalabs/gaudi2/gaudi2P.h @@ -19,8 +19,6 @@ #define GAUDI2_LINUX_FW_FILE "habanalabs/gaudi2/gaudi2-fit.itb" #define GAUDI2_BOOT_FIT_FILE "habanalabs/gaudi2/gaudi2-boot-fit.itb" -#define MMU_PAGE_TABLES_INITIAL_SIZE 0x10000000 /* 256MB */ - #define GAUDI2_CPU_TIMEOUT_USEC 30000000 /* 30s */ #define NUMBER_OF_PDMA_QUEUES 2 @@ -109,13 +107,11 @@ /* DRAM Memory Map */ #define CPU_FW_IMAGE_SIZE 0x10000000 /* 256MB */ - -/* This define should be used only when working in a debug mode without dram. - * When working with dram, the driver size will be calculated dynamically. - */ -#define NIC_DEFAULT_DRV_SIZE 0x20000000 /* 512MB */ - #define CPU_FW_IMAGE_ADDR DRAM_PHYS_BASE +#define PMMU_PAGE_TABLES_SIZE 0x10000000 /* 256MB */ +#define EDMA_PQS_SIZE SZ_2M +#define EDMA_SCRATCHPAD_SIZE SZ_1M +#define HMMU_PAGE_TABLES_SIZE SZ_1M #define NIC_NUMBER_OF_PORTS NIC_NUMBER_OF_ENGINES @@ -241,9 +237,8 @@ #define GAUDI2_SOB_INCREMENT_BY_ONE (FIELD_PREP(DCORE0_SYNC_MNGR_OBJS_SOB_OBJ_VAL_MASK, 1) | \ FIELD_PREP(DCORE0_SYNC_MNGR_OBJS_SOB_OBJ_INC_MASK, 1)) -#define GAUDI2_NUM_TESTED_QS (GAUDI2_QUEUE_ID_CPU_PQ - GAUDI2_QUEUE_ID_PDMA_0_0) +#define GAUDI2_NUM_TESTED_QS (GAUDI2_QUEUE_ID_CPU_PQ - GAUDI2_QUEUE_ID_PDMA_0_0) -#define GAUDI2_NUM_OF_GLBL_ERR_CAUSE 8 enum gaudi2_reserved_sob_id { GAUDI2_RESERVED_SOB_CS_COMPLETION_FIRST, @@ -389,7 +384,7 @@ enum gaudi2_edma_id { /* User interrupt count is aligned with HW CQ count. * We have 64 CQ's per dcore, CQ0 in dcore 0 is reserved for legacy mode */ -#define GAUDI2_NUM_USER_INTERRUPTS 255 +#define GAUDI2_NUM_USER_INTERRUPTS 64 #define GAUDI2_NUM_RESERVED_INTERRUPTS 1 #define GAUDI2_TOTAL_USER_INTERRUPTS (GAUDI2_NUM_USER_INTERRUPTS + GAUDI2_NUM_RESERVED_INTERRUPTS) @@ -421,11 +416,11 @@ enum gaudi2_irq_num { GAUDI2_IRQ_NUM_NIC_PORT_LAST = (GAUDI2_IRQ_NUM_NIC_PORT_FIRST + NIC_NUMBER_OF_PORTS - 1), GAUDI2_IRQ_NUM_TPC_ASSERT, GAUDI2_IRQ_NUM_EQ_ERROR, + GAUDI2_IRQ_NUM_USER_FIRST, + GAUDI2_IRQ_NUM_USER_LAST = (GAUDI2_IRQ_NUM_USER_FIRST + GAUDI2_NUM_USER_INTERRUPTS - 1), GAUDI2_IRQ_NUM_RESERVED_FIRST, - GAUDI2_IRQ_NUM_RESERVED_LAST = (GAUDI2_MSIX_ENTRIES - GAUDI2_TOTAL_USER_INTERRUPTS - 1), + GAUDI2_IRQ_NUM_RESERVED_LAST = (GAUDI2_MSIX_ENTRIES - GAUDI2_NUM_RESERVED_INTERRUPTS - 1), GAUDI2_IRQ_NUM_UNEXPECTED_ERROR = RESERVED_MSIX_UNEXPECTED_USER_ERROR_INTERRUPT, - GAUDI2_IRQ_NUM_USER_FIRST = GAUDI2_IRQ_NUM_UNEXPECTED_ERROR + 1, - GAUDI2_IRQ_NUM_USER_LAST = (GAUDI2_IRQ_NUM_USER_FIRST + GAUDI2_NUM_USER_INTERRUPTS - 1), GAUDI2_IRQ_NUM_LAST = (GAUDI2_MSIX_ENTRIES - 1) }; diff --git a/drivers/accel/habanalabs/gaudi2/gaudi2_security.c b/drivers/accel/habanalabs/gaudi2/gaudi2_security.c index 34bf80c5a44b..307ccb912ccd 100644 --- a/drivers/accel/habanalabs/gaudi2/gaudi2_security.c +++ b/drivers/accel/habanalabs/gaudi2/gaudi2_security.c @@ -479,6 +479,7 @@ static const u32 gaudi2_pb_dcr0_edma0_unsecured_regs[] = { mmDCORE0_EDMA0_CORE_CTX_TE_NUMROWS, mmDCORE0_EDMA0_CORE_CTX_IDX, mmDCORE0_EDMA0_CORE_CTX_IDX_INC, + mmDCORE0_EDMA0_CORE_WR_COMP_MAX_OUTSTAND, mmDCORE0_EDMA0_CORE_RD_LBW_RATE_LIM_CFG, mmDCORE0_EDMA0_QM_CQ_CFG0_0, mmDCORE0_EDMA0_QM_CQ_CFG0_1, diff --git a/drivers/accel/habanalabs/goya/goya.c b/drivers/accel/habanalabs/goya/goya.c index 1322cb330c57..84768e306269 100644 --- a/drivers/accel/habanalabs/goya/goya.c +++ b/drivers/accel/habanalabs/goya/goya.c @@ -413,8 +413,6 @@ int goya_set_fixed_properties(struct hl_device *hdev) else prop->mmu_pgt_size = MMU_PAGE_TABLES_SIZE; prop->mmu_pte_size = HL_PTE_SIZE; - prop->mmu_hop_table_size = HOP_TABLE_SIZE_512_PTE; - prop->mmu_hop0_tables_total_size = HOP0_512_PTE_TABLES_TOTAL_SIZE; prop->dram_page_size = PAGE_SIZE_2MB; prop->device_mem_alloc_default_page_size = prop->dram_page_size; prop->dram_supports_virtual_memory = true; @@ -435,8 +433,8 @@ int goya_set_fixed_properties(struct hl_device *hdev) prop->dmmu.num_hops = MMU_ARCH_5_HOPS; prop->dmmu.last_mask = LAST_MASK; /* TODO: will be duplicated until implementing per-MMU props */ - prop->dmmu.hop_table_size = prop->mmu_hop_table_size; - prop->dmmu.hop0_tables_total_size = prop->mmu_hop0_tables_total_size; + prop->dmmu.hop_table_size = HOP_TABLE_SIZE_512_PTE; + prop->dmmu.hop0_tables_total_size = HOP0_512_PTE_TABLES_TOTAL_SIZE; /* shifts and masks are the same in PMMU and DMMU */ memcpy(&prop->pmmu, &prop->dmmu, sizeof(prop->dmmu)); @@ -446,8 +444,8 @@ int goya_set_fixed_properties(struct hl_device *hdev) prop->pmmu.num_hops = MMU_ARCH_5_HOPS; prop->pmmu.last_mask = LAST_MASK; /* TODO: will be duplicated until implementing per-MMU props */ - prop->pmmu.hop_table_size = prop->mmu_hop_table_size; - prop->pmmu.hop0_tables_total_size = prop->mmu_hop0_tables_total_size; + prop->pmmu.hop_table_size = HOP_TABLE_SIZE_512_PTE; + prop->pmmu.hop0_tables_total_size = HOP0_512_PTE_TABLES_TOTAL_SIZE; /* PMMU and HPMMU are the same except of page size */ memcpy(&prop->pmmu_huge, &prop->pmmu, sizeof(prop->pmmu)); @@ -895,11 +893,8 @@ int goya_late_init(struct hl_device *hdev) WREG32(mmMMU_LOG2_DDR_SIZE, ilog2(prop->dram_size)); rc = hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_ENABLE_PCI_ACCESS, 0x0); - if (rc) { - dev_err(hdev->dev, - "Failed to enable PCI access from CPU %d\n", rc); + if (rc) return rc; - } /* force setting to low frequency */ goya->curr_pll_profile = PLL_LOW; @@ -2678,7 +2673,7 @@ int goya_mmu_init(struct hl_device *hdev) for (i = 0 ; i < prop->max_asid ; i++) { hop0_addr = prop->mmu_pgt_addr + - (i * prop->mmu_hop_table_size); + (i * prop->dmmu.hop_table_size); rc = goya_mmu_update_asid_hop0_addr(hdev, i, hop0_addr); if (rc) { @@ -2866,13 +2861,7 @@ static int goya_hw_fini(struct hl_device *hdev, bool hard_reset, bool fw_reset) int goya_suspend(struct hl_device *hdev) { - int rc; - - rc = hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0); - if (rc) - dev_err(hdev->dev, "Failed to disable PCI access from CPU\n"); - - return rc; + return hl_fw_send_pci_access_msg(hdev, CPUCP_PACKET_DISABLE_PCI_ACCESS, 0x0); } int goya_resume(struct hl_device *hdev) diff --git a/drivers/accel/habanalabs/goya/goya_coresight.c b/drivers/accel/habanalabs/goya/goya_coresight.c index 41cae5fd843b..3827ea4c02f7 100644 --- a/drivers/accel/habanalabs/goya/goya_coresight.c +++ b/drivers/accel/habanalabs/goya/goya_coresight.c @@ -576,7 +576,6 @@ static int goya_config_spmu(struct hl_device *hdev, struct hl_debug_params *params) { u64 base_reg; - struct hl_debug_params_spmu *input = params->input; u64 *output; u32 output_arr_len; u32 events_num; @@ -592,7 +591,7 @@ static int goya_config_spmu(struct hl_device *hdev, base_reg = debug_spmu_regs[params->reg_idx] - CFG_BASE; if (params->enable) { - input = params->input; + struct hl_debug_params_spmu *input = params->input; if (!input) return -EINVAL; diff --git a/drivers/accel/habanalabs/include/gaudi2/gaudi2.h b/drivers/accel/habanalabs/include/gaudi2/gaudi2.h index 0231d6c55b4a..753d46a2836b 100644 --- a/drivers/accel/habanalabs/include/gaudi2/gaudi2.h +++ b/drivers/accel/habanalabs/include/gaudi2/gaudi2.h @@ -63,9 +63,9 @@ #define RESERVED_VA_RANGE_FOR_ARC_ON_HOST_HPAGE_START 0xFFF0F80000000000ull #define RESERVED_VA_RANGE_FOR_ARC_ON_HOST_HPAGE_END 0xFFF0FFFFFFFFFFFFull -#define RESERVED_MSIX_UNEXPECTED_USER_ERROR_INTERRUPT 256 +#define RESERVED_MSIX_UNEXPECTED_USER_ERROR_INTERRUPT 127 -#define GAUDI2_MSIX_ENTRIES 512 +#define GAUDI2_MSIX_ENTRIES 128 #define QMAN_PQ_ENTRY_SIZE 16 /* Bytes */ diff --git a/drivers/accel/habanalabs/include/gaudi2/gaudi2_async_ids_map_extended.h b/drivers/accel/habanalabs/include/gaudi2/gaudi2_async_ids_map_extended.h index b2dbe1f64430..82d639990cca 100644 --- a/drivers/accel/habanalabs/include/gaudi2/gaudi2_async_ids_map_extended.h +++ b/drivers/accel/habanalabs/include/gaudi2/gaudi2_async_ids_map_extended.h @@ -330,9 +330,9 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { { .fc_id = 149, .cpu_id = 48, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "EDMA7_ECC_SERR" }, { .fc_id = 150, .cpu_id = 48, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, - .name = "HDMA4_ECC_SERR" }, + .name = "EDMA4_ECC_SERR" }, { .fc_id = 151, .cpu_id = 48, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, - .name = "HDMA5_ECC_SERR" }, + .name = "EDMA5_ECC_SERR" }, { .fc_id = 152, .cpu_id = 49, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "EDMA2_ECC_DERR" }, { .fc_id = 153, .cpu_id = 49, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, @@ -856,55 +856,55 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { { .fc_id = 412, .cpu_id = 84, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "PCIE_ADDR_DEC_ERR" }, { .fc_id = 413, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC0_AXI_ERR_RSP" }, + .name = "DCORE0_TPC0_AXI_ERR_RSP" }, { .fc_id = 414, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC1_AXI_ERR_RSP" }, + .name = "DCORE0_TPC1_AXI_ERR_RSP" }, { .fc_id = 415, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC2_AXI_ERR_RSP" }, + .name = "DCORE0_TPC2_AXI_ERR_RSP" }, { .fc_id = 416, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC3_AXI_ERR_RSP" }, + .name = "DCORE0_TPC3_AXI_ERR_RSP" }, { .fc_id = 417, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC4_AXI_ERR_RSP" }, + .name = "DCORE0_TPC4_AXI_ERR_RSP" }, { .fc_id = 418, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC5_AXI_ERR_RSP" }, + .name = "DCORE0_TPC5_AXI_ERR_RSP" }, { .fc_id = 419, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC6_AXI_ERR_RSP" }, + .name = "DCORE1_TPC0_AXI_ERR_RSP" }, { .fc_id = 420, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC7_AXI_ERR_RSP" }, + .name = "DCORE1_TPC1_AXI_ERR_RSP" }, { .fc_id = 421, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC8_AXI_ERR_RSP" }, + .name = "DCORE1_TPC2_AXI_ERR_RSP" }, { .fc_id = 422, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC9_AXI_ERR_RSP" }, + .name = "DCORE1_TPC3_AXI_ERR_RSP" }, { .fc_id = 423, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC10_AXI_ERR_RSP" }, + .name = "DCORE1_TPC4_AXI_ERR_RSP" }, { .fc_id = 424, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC11_AXI_ERR_RSP" }, + .name = "DCORE1_TPC5_AXI_ERR_RSP" }, { .fc_id = 425, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC12_AXI_ERR_RSP" }, + .name = "DCORE2_TPC0_AXI_ERR_RSP" }, { .fc_id = 426, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC13_AXI_ERR_RSP" }, + .name = "DCORE2_TPC1_AXI_ERR_RSP" }, { .fc_id = 427, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC14_AXI_ERR_RSP" }, + .name = "DCORE2_TPC2_AXI_ERR_RSP" }, { .fc_id = 428, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC15_AXI_ERR_RSP" }, + .name = "DCORE2_TPC3_AXI_ERR_RSP" }, { .fc_id = 429, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC16_AXI_ERR_RSP" }, + .name = "DCORE2_TPC4_AXI_ERR_RSP" }, { .fc_id = 430, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC17_AXI_ERR_RSP" }, + .name = "DCORE2_TPC5_AXI_ERR_RSP" }, { .fc_id = 431, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC18_AXI_ERR_RSP" }, + .name = "DCORE3_TPC0_AXI_ERR_RSP" }, { .fc_id = 432, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC19_AXI_ERR_RSP" }, + .name = "DCORE3_TPC1_AXI_ERR_RSP" }, { .fc_id = 433, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC20_AXI_ERR_RSP" }, + .name = "DCORE3_TPC2_AXI_ERR_RSP" }, { .fc_id = 434, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC21_AXI_ERR_RSP" }, + .name = "DCORE3_TPC3_AXI_ERR_RSP" }, { .fc_id = 435, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC22_AXI_ERR_RSP" }, + .name = "DCORE3_TPC4_AXI_ERR_RSP" }, { .fc_id = 436, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC23_AXI_ERR_RSP" }, + .name = "DCORE3_TPC5_AXI_ERR_RSP" }, { .fc_id = 437, .cpu_id = 85, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC24_AXI_ERR_RSP" }, + .name = "DCORE4_TPC0_AXI_ERR_RSP" }, { .fc_id = 438, .cpu_id = 86, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "AXI_ECC" }, { .fc_id = 439, .cpu_id = 87, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, @@ -965,73 +965,73 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { .name = "MME3_CTRL_AXI_ERROR_RESPONSE" }, { .fc_id = 467, .cpu_id = 91, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, .name = "MME3_QMAN_SW_ERROR" }, - { .fc_id = 468, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 468, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "PSOC_MME_PLL_LOCK_ERR" }, - { .fc_id = 469, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 469, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "PSOC_CPU_PLL_LOCK_ERR" }, - { .fc_id = 470, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 470, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_TPC_PLL_LOCK_ERR" }, - { .fc_id = 471, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 471, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_NIC_PLL_LOCK_ERR" }, - { .fc_id = 472, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 472, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_XBAR_MMU_PLL_LOCK_ERR" }, - { .fc_id = 473, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 473, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_XBAR_DMA_PLL_LOCK_ERR" }, - { .fc_id = 474, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 474, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_XBAR_IF_PLL_LOCK_ERR" }, - { .fc_id = 475, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 475, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_XBAR_BANK_PLL_LOCK_ERR" }, - { .fc_id = 476, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 476, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_XBAR_MMU_PLL_LOCK_ERR" }, - { .fc_id = 477, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 477, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_XBAR_DMA_PLL_LOCK_ERR" }, - { .fc_id = 478, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 478, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_XBAR_IF_PLL_LOCK_ERR" }, - { .fc_id = 479, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 479, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_XBAR_MESH_PLL_LOCK_ERR" }, - { .fc_id = 480, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 480, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_TPC_PLL_LOCK_ERR" }, - { .fc_id = 481, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 481, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_NIC_PLL_LOCK_ERR" }, - { .fc_id = 482, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 482, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "PMMU_MME_PLL_LOCK_ERR" }, - { .fc_id = 483, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 483, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_TPC_PLL_LOCK_ERR" }, - { .fc_id = 484, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 484, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_PCI_PLL_LOCK_ERR" }, - { .fc_id = 485, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 485, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_XBAR_MMU_PLL_LOCK_ERR" }, - { .fc_id = 486, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 486, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_XBAR_DMA_PLL_LOCK_ERR" }, - { .fc_id = 487, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 487, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_XBAR_IF_PLL_LOCK_ERR" }, - { .fc_id = 488, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 488, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_XBAR_MESH_PLL_LOCK_ERR" }, - { .fc_id = 489, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 489, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_XBAR_MMU_PLL_LOCK_ERR" }, - { .fc_id = 490, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 490, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_XBAR_DMA_PLL_LOCK_ERR" }, - { .fc_id = 491, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 491, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_XBAR_IF_PLL_LOCK_ERR" }, - { .fc_id = 492, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 492, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_XBAR_BANK_PLL_LOCK_ERR" }, - { .fc_id = 493, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 493, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_TPC_PLL_LOCK_ERR" }, - { .fc_id = 494, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 494, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "PSOC_VID_PLL_LOCK_ERR" }, - { .fc_id = 495, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 495, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "PMMU_VID_PLL_LOCK_ERR" }, - { .fc_id = 496, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 496, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE3_HBM_PLL_LOCK_ERR" }, - { .fc_id = 497, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 497, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_XBAR_HBM_PLL_LOCK_ERR" }, - { .fc_id = 498, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 498, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE1_HBM_PLL_LOCK_ERR" }, - { .fc_id = 499, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 499, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE0_HBM_PLL_LOCK_ERR" }, - { .fc_id = 500, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 500, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_XBAR_HBM_PLL_LOCK_ERR" }, - { .fc_id = 501, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + { .fc_id = 501, .cpu_id = 92, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "DCORE2_HBM_PLL_LOCK_ERR" }, { .fc_id = 502, .cpu_id = 93, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_HARD, .name = "CPU_AXI_ERR_RSP" }, @@ -1298,103 +1298,103 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { { .fc_id = 633, .cpu_id = 130, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC0_BMON_SPMU" }, { .fc_id = 634, .cpu_id = 131, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC0_KERNEL_ERR" }, + .name = "DCORE0_TPC0_KERNEL_ERR" }, { .fc_id = 635, .cpu_id = 132, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC1_BMON_SPMU" }, { .fc_id = 636, .cpu_id = 133, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC1_KERNEL_ERR" }, + .name = "DCORE0_TPC1_KERNEL_ERR" }, { .fc_id = 637, .cpu_id = 134, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC2_BMON_SPMU" }, { .fc_id = 638, .cpu_id = 135, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC2_KERNEL_ERR" }, + .name = "DCORE0_TPC2_KERNEL_ERR" }, { .fc_id = 639, .cpu_id = 136, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC3_BMON_SPMU" }, { .fc_id = 640, .cpu_id = 137, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC3_KERNEL_ERR" }, + .name = "DCORE0_TPC3_KERNEL_ERR" }, { .fc_id = 641, .cpu_id = 138, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC4_BMON_SPMU" }, { .fc_id = 642, .cpu_id = 139, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC4_KERNEL_ERR" }, + .name = "DCORE0_TPC4_KERNEL_ERR" }, { .fc_id = 643, .cpu_id = 140, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC5_BMON_SPMU" }, { .fc_id = 644, .cpu_id = 141, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC5_KERNEL_ERR" }, + .name = "DCORE0_TPC5_KERNEL_ERR" }, { .fc_id = 645, .cpu_id = 150, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC6_BMON_SPMU" }, { .fc_id = 646, .cpu_id = 151, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC6_KERNEL_ERR" }, + .name = "DCORE1_TPC0_KERNEL_ERR" }, { .fc_id = 647, .cpu_id = 152, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC7_BMON_SPMU" }, { .fc_id = 648, .cpu_id = 153, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC7_KERNEL_ERR" }, + .name = "DCORE1_TPC1_KERNEL_ERR" }, { .fc_id = 649, .cpu_id = 146, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC8_BMON_SPMU" }, { .fc_id = 650, .cpu_id = 147, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC8_KERNEL_ERR" }, + .name = "DCORE1_TPC2_KERNEL_ERR" }, { .fc_id = 651, .cpu_id = 148, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC9_BMON_SPMU" }, { .fc_id = 652, .cpu_id = 149, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC9_KERNEL_ERR" }, + .name = "DCORE1_TPC3_KERNEL_ERR" }, { .fc_id = 653, .cpu_id = 142, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC10_BMON_SPMU" }, { .fc_id = 654, .cpu_id = 143, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC10_KERNEL_ERR" }, + .name = "DCORE1_TPC4_KERNEL_ERR" }, { .fc_id = 655, .cpu_id = 144, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC11_BMON_SPMU" }, { .fc_id = 656, .cpu_id = 145, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC11_KERNEL_ERR" }, + .name = "DCORE1_TPC5_KERNEL_ERR" }, { .fc_id = 657, .cpu_id = 162, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC12_BMON_SPMU" }, { .fc_id = 658, .cpu_id = 163, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC12_KERNEL_ERR" }, + .name = "DCORE2_TPC0_KERNEL_ERR" }, { .fc_id = 659, .cpu_id = 164, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC13_BMON_SPMU" }, { .fc_id = 660, .cpu_id = 165, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC13_KERNEL_ERR" }, + .name = "DCORE2_TPC1_KERNEL_ERR" }, { .fc_id = 661, .cpu_id = 158, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC14_BMON_SPMU" }, { .fc_id = 662, .cpu_id = 159, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC14_KERNEL_ERR" }, + .name = "DCORE2_TPC2_KERNEL_ERR" }, { .fc_id = 663, .cpu_id = 160, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC15_BMON_SPMU" }, { .fc_id = 664, .cpu_id = 161, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC15_KERNEL_ERR" }, + .name = "DCORE2_TPC3_KERNEL_ERR" }, { .fc_id = 665, .cpu_id = 154, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC16_BMON_SPMU" }, { .fc_id = 666, .cpu_id = 155, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC16_KERNEL_ERR" }, + .name = "DCORE2_TPC4_KERNEL_ERR" }, { .fc_id = 667, .cpu_id = 156, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC17_BMON_SPMU" }, { .fc_id = 668, .cpu_id = 157, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC17_KERNEL_ERR" }, + .name = "DCORE2_TPC5_KERNEL_ERR" }, { .fc_id = 669, .cpu_id = 166, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC18_BMON_SPMU" }, { .fc_id = 670, .cpu_id = 167, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC18_KERNEL_ERR" }, + .name = "DCORE3_TPC0_KERNEL_ERR" }, { .fc_id = 671, .cpu_id = 168, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC19_BMON_SPMU" }, { .fc_id = 672, .cpu_id = 169, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC19_KERNEL_ERR" }, + .name = "DCORE3_TPC1_KERNEL_ERR" }, { .fc_id = 673, .cpu_id = 170, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC20_BMON_SPMU" }, { .fc_id = 674, .cpu_id = 171, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC20_KERNEL_ERR" }, + .name = "DCORE3_TPC2_KERNEL_ERR" }, { .fc_id = 675, .cpu_id = 172, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC21_BMON_SPMU" }, { .fc_id = 676, .cpu_id = 173, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC21_KERNEL_ERR" }, + .name = "DCORE3_TPC3_KERNEL_ERR" }, { .fc_id = 677, .cpu_id = 174, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC22_BMON_SPMU" }, { .fc_id = 678, .cpu_id = 175, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC22_KERNEL_ERR" }, + .name = "DCORE3_TPC4_KERNEL_ERR" }, { .fc_id = 679, .cpu_id = 176, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC23_BMON_SPMU" }, { .fc_id = 680, .cpu_id = 177, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC23_KERNEL_ERR" }, + .name = "DCORE3_TPC5_KERNEL_ERR" }, { .fc_id = 681, .cpu_id = 178, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "TPC24_BMON_SPMU" }, { .fc_id = 682, .cpu_id = 179, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC24_KERNEL_ERR" }, + .name = "DCORE4_TPC0_KERNEL_ERR" }, { .fc_id = 683, .cpu_id = 180, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "" }, { .fc_id = 684, .cpu_id = 180, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, @@ -1827,8 +1827,8 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { .name = "DEC0_BMON_SPMU" }, { .fc_id = 898, .cpu_id = 330, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, .name = "DEC1_SPI" }, - { .fc_id = 899, .cpu_id = 330, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "DEC1_SPI" }, + { .fc_id = 899, .cpu_id = 330, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, + .name = "DEC1_BMON_SPMU" }, { .fc_id = 900, .cpu_id = 331, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_COMPUTE, .name = "DEC2_SPI" }, { .fc_id = 901, .cpu_id = 331, .valid = 1, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, @@ -2377,8 +2377,8 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { .name = "" }, { .fc_id = 1173, .cpu_id = 479, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "" }, - { .fc_id = 1174, .cpu_id = 480, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, - .name = "" }, + { .fc_id = 1174, .cpu_id = 480, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, + .name = "PSOC_DMA_QM" }, { .fc_id = 1175, .cpu_id = 481, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "" }, { .fc_id = 1176, .cpu_id = 482, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, @@ -2442,55 +2442,55 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { { .fc_id = 1205, .cpu_id = 511, .valid = 0, .msg = 0, .reset = EVENT_RESET_TYPE_NONE, .name = "" }, { .fc_id = 1206, .cpu_id = 512, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC0_QM" }, + .name = "DCORE0_TPC0_QM" }, { .fc_id = 1207, .cpu_id = 513, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC1_QM" }, + .name = "DCORE0_TPC1_QM" }, { .fc_id = 1208, .cpu_id = 514, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC2_QM" }, + .name = "DCORE0_TPC2_QM" }, { .fc_id = 1209, .cpu_id = 515, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC3_QM" }, + .name = "DCORE0_TPC3_QM" }, { .fc_id = 1210, .cpu_id = 516, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC4_QM" }, + .name = "DCORE0_TPC4_QM" }, { .fc_id = 1211, .cpu_id = 517, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC5_QM" }, + .name = "DCORE0_TPC5_QM" }, { .fc_id = 1212, .cpu_id = 518, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC6_QM" }, + .name = "DCORE1_TPC0_QM" }, { .fc_id = 1213, .cpu_id = 519, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC7_QM" }, + .name = "DCORE1_TPC1_QM" }, { .fc_id = 1214, .cpu_id = 520, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC8_QM" }, + .name = "DCORE1_TPC2_QM" }, { .fc_id = 1215, .cpu_id = 521, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC9_QM" }, + .name = "DCORE1_TPC3_QM" }, { .fc_id = 1216, .cpu_id = 522, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC10_QM" }, + .name = "DCORE1_TPC4_QM" }, { .fc_id = 1217, .cpu_id = 523, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC11_QM" }, + .name = "DCORE1_TPC5_QM" }, { .fc_id = 1218, .cpu_id = 524, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC12_QM" }, + .name = "DCORE2_TPC0_QM" }, { .fc_id = 1219, .cpu_id = 525, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC13_QM" }, + .name = "DCORE2_TPC1_QM" }, { .fc_id = 1220, .cpu_id = 526, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC14_QM" }, + .name = "DCORE2_TPC2_QM" }, { .fc_id = 1221, .cpu_id = 527, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC15_QM" }, + .name = "DCORE2_TPC3_QM" }, { .fc_id = 1222, .cpu_id = 528, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC16_QM" }, + .name = "DCORE2_TPC4_QM" }, { .fc_id = 1223, .cpu_id = 529, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC17_QM" }, + .name = "DCORE2_TPC5_QM" }, { .fc_id = 1224, .cpu_id = 530, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC18_QM" }, + .name = "DCORE3_TPC0_QM" }, { .fc_id = 1225, .cpu_id = 531, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC19_QM" }, + .name = "DCORE3_TPC1_QM" }, { .fc_id = 1226, .cpu_id = 532, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC20_QM" }, + .name = "DCORE3_TPC2_QM" }, { .fc_id = 1227, .cpu_id = 533, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC21_QM" }, + .name = "DCORE3_TPC3_QM" }, { .fc_id = 1228, .cpu_id = 534, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC22_QM" }, + .name = "DCORE3_TPC4_QM" }, { .fc_id = 1229, .cpu_id = 535, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC23_QM" }, + .name = "DCORE3_TPC5_QM" }, { .fc_id = 1230, .cpu_id = 536, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, - .name = "TPC24_QM" }, + .name = "DCORE4_TPC0_QM" }, { .fc_id = 1231, .cpu_id = 537, .valid = 0, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, .name = "" }, { .fc_id = 1232, .cpu_id = 538, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_COMPUTE, @@ -2674,19 +2674,19 @@ static struct gaudi2_async_events_ids_map gaudi2_irq_map_table[] = { { .fc_id = 1321, .cpu_id = 627, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_HARD, .name = "DEV_RESET_REQ" }, { .fc_id = 1322, .cpu_id = 628, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_PWR_BRK_ENTRY" }, + .name = "PWR_BRK_ENTRY" }, { .fc_id = 1323, .cpu_id = 629, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_PWR_BRK_EXT" }, + .name = "PWR_BRK_EXT" }, { .fc_id = 1324, .cpu_id = 630, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_PWR_RD_MODE0" }, + .name = "PWR_RD_MODE0" }, { .fc_id = 1325, .cpu_id = 631, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_PWR_RD_MODE1" }, + .name = "PWR_RD_MODE1" }, { .fc_id = 1326, .cpu_id = 632, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_PWR_RD_MODE2" }, + .name = "PWR_RD_MODE2" }, { .fc_id = 1327, .cpu_id = 633, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_PWR_RD_MODE3" }, + .name = "PWR_RD_MODE3" }, { .fc_id = 1328, .cpu_id = 634, .valid = 1, .msg = 1, .reset = EVENT_RESET_TYPE_NONE, - .name = "ARC_EQ_HEARTBEAT" }, + .name = "EQ_HEARTBEAT" }, }; #endif /* __GAUDI2_ASYNC_IDS_MAP_EVENTS_EXT_H_ */ diff --git a/drivers/accel/habanalabs/include/gaudi2/gaudi2_fw_if.h b/drivers/accel/habanalabs/include/gaudi2/gaudi2_fw_if.h index 18ca147b1c86..6ea936c9594e 100644 --- a/drivers/accel/habanalabs/include/gaudi2/gaudi2_fw_if.h +++ b/drivers/accel/habanalabs/include/gaudi2/gaudi2_fw_if.h @@ -45,6 +45,13 @@ #define GAUDI2_ARM_RX_MB_OFFSET (GAUDI2_ARM_RX_MB_ADDR - \ GAUDI2_SP_SRAM_BASE_ADDR) +#define POWER_MODE_LEVELS { \ + 150000, /* 00 */ \ + 250000, /* 01 */ \ + 400000, /* 10 */ \ + /* 11: Normal mode */ \ +} + enum gaudi2_fw_status { GAUDI2_PID_STATUS_UP = 0x1, /* PID on ARC0 is up */ GAUDI2_ARM_STATUS_UP = 0x2, /* ARM Linux Boot complete */ @@ -52,26 +59,6 @@ enum gaudi2_fw_status { GAUDI2_STATUS_LAST = 0xFF }; -struct gaudi2_cold_rst_data { - union { - struct { - u32 recovery_flag: 1; - u32 validation_flag: 1; - u32 efuse_read_flag: 1; - u32 spsram_init_done : 1; - u32 fake_security_enable : 1; - u32 fake_sig_validation_en : 1; - u32 bist_skip_enable : 1; - u32 reserved1 : 1; - u32 fake_bis_compliant : 1; - u32 wd_rst_cause_arm : 1; - u32 wd_rst_cause_arcpid : 1; - u32 reserved : 21; - }; - __le32 data; - }; -}; - enum gaudi2_rst_src { HL_COLD_RST = 1, HL_MANUAL_RST = 2, diff --git a/drivers/accel/habanalabs/include/gaudi2/gaudi2_reg_map.h b/drivers/accel/habanalabs/include/gaudi2/gaudi2_reg_map.h index f3eaeb6d9b7e..1e9c056e437d 100644 --- a/drivers/accel/habanalabs/include/gaudi2/gaudi2_reg_map.h +++ b/drivers/accel/habanalabs/include/gaudi2/gaudi2_reg_map.h @@ -58,4 +58,12 @@ #define mmWD_GPIO_DATAOUT_REG mmPSOC_GPIO3_DATAOUT #define mmSTM_PROFILER_SPE_REG mmPSOC_STM_STMSPER +/* Registers below are used to pass the boot_if data between ARM and ARC1 */ +#define mmARM_MSG_BOOT_ERR_SET mmCPU_IF_SPECIAL_GLBL_SPARE_0 +#define mmARM_MSG_BOOT_ERR_CLR mmCPU_IF_SPECIAL_GLBL_SPARE_1 +#define mmARM_MSG_BOOT_DEV_STS_SET mmCPU_IF_SPECIAL_GLBL_SPARE_2 +#define mmARM_MSG_BOOT_DEV_STS_CLR mmCPU_IF_SPECIAL_GLBL_SPARE_3 +#define mmMGMT_MSG_BOOT_ERR mmCPU_MSTR_IF_SPECIAL_GLBL_SPARE_0 +#define mmMGMT_MSG_BOOT_DEV_STS mmCPU_MSTR_IF_SPECIAL_GLBL_SPARE_1 + #endif /* GAUDI2_REG_MAP_H_ */ diff --git a/drivers/accel/habanalabs/include/hw_ip/mmu/mmu_general.h b/drivers/accel/habanalabs/include/hw_ip/mmu/mmu_general.h index d408feecd483..b4a5e95be354 100644 --- a/drivers/accel/habanalabs/include/hw_ip/mmu/mmu_general.h +++ b/drivers/accel/habanalabs/include/hw_ip/mmu/mmu_general.h @@ -26,6 +26,8 @@ #define LAST_MASK 0x0000000000800ull #define FLAGS_MASK 0x0000000000FFFull +#define MMU_ARCH_3_HOPS 3 +#define MMU_ARCH_4_HOPS 4 #define MMU_ARCH_5_HOPS 5 #define MMU_ARCH_6_HOPS 6 diff --git a/drivers/accel/habanalabs/include/hw_ip/pci/pci_general.h b/drivers/accel/habanalabs/include/hw_ip/pci/pci_general.h index 4f951cada077..a75faa00197f 100644 --- a/drivers/accel/habanalabs/include/hw_ip/pci/pci_general.h +++ b/drivers/accel/habanalabs/include/hw_ip/pci/pci_general.h @@ -25,7 +25,8 @@ enum hl_revision_id { REV_ID_INVALID = 0x00, REV_ID_A = 0x01, REV_ID_B = 0x02, - REV_ID_C = 0x03 + REV_ID_C = 0x03, + REV_ID_D = 0x04 }; #endif /* INCLUDE_PCI_GENERAL_H_ */ |