Merge tag 'drm-misc-next-2025-10-02' of https://gitlab.freedesktop.org/drm/misc/kernel into drm-next

drm-misc-next for v6.19:

UAPI Changes:

Cross-subsystem Changes:
-  fbcon cleanups.
- Make drivers depend on FB_TILEBLITTING instead of selecting it,
  and hide FB_MODE_HELPERS.

Core Changes:
- More preparations for rust.
- Throttle dirty worker with vblank
- Use drm_for_each_bridge_in_chain_scoped in drm's bridge code and
  assorted fixes.
- Ensure drm_client_modeset tests are enabled in UML.
- Rename ttm_bo_put to ttm_bo_fini, as a further step in removing the
  TTM bo refcount.
- Add POST_LT_ADJ_REQ training sequence.
- Show list of removed but still allocated bridges.
- Add a simulated vblank interrupt for hardware without it,
  and add some helpers to use them in vkms and hypervdrm.

Driver Changes:
- Assorted small fixes, cleanups and updates to host1x, tegra,
  panthor,   amdxdna, gud, vc4, ssd130x, ivpu, panfrost, panthor,
  sysfb, bridge/sn65dsi86, solomon, ast, tidss.
- Convert drivers from using .round_rate() to .determine_rate()
- Add support for KD116N3730A07/A12, chromebook mt8189, JT101TM023,
  LQ079L1SX01, raspberrypi 5" panels.
- Improve reclocking on tegra186+ with nouveau.
- Improve runtime pm in amdxdna.
- Add support for HTX_PAI in imx.
- Use a helper to calculate dumb buffer sizes in most drivers.

Signed-off-by: Simona Vetter <simona.vetter@ffwll.ch>
From: Maarten Lankhorst <maarten.lankhorst@linux.intel.com>
Link: https://lore.kernel.org/r/b412fb91-8545-466a-8102-d89c0f2758a7@linux.intel.com

28 files changed, 1051 insertions, 581 deletions

diff --git a/drivers/accel/amdxdna/Makefile b/drivers/accel/amdxdna/Makefile
index 6797dac65efa..6344aaf523fa 100644
--- a/drivers/accel/amdxdna/Makefile
+++ b/drivers/accel/amdxdna/Makefile
@@ -14,6 +14,7 @@ amdxdna-y := \
 	amdxdna_mailbox.o \
 	amdxdna_mailbox_helper.o \
 	amdxdna_pci_drv.o \
+	amdxdna_pm.o \
 	amdxdna_sysfs.o \
 	amdxdna_ubuf.o \
 	npu1_regs.o \
diff --git a/drivers/accel/amdxdna/aie2_ctx.c b/drivers/accel/amdxdna/aie2_ctx.c
index e9f9b1fa5dc1..691fdb3b008f 100644
--- a/drivers/accel/amdxdna/aie2_ctx.c
+++ b/drivers/accel/amdxdna/aie2_ctx.c
@@ -21,6 +21,7 @@
 #include "amdxdna_gem.h"
 #include "amdxdna_mailbox.h"
 #include "amdxdna_pci_drv.h"
+#include "amdxdna_pm.h"
 
 static bool force_cmdlist;
 module_param(force_cmdlist, bool, 0600);
@@ -88,7 +89,7 @@ static int aie2_hwctx_restart(struct amdxdna_dev *xdna, struct amdxdna_hwctx *hw
 		goto out;
 	}
 
-	ret = aie2_config_cu(hwctx);
+	ret = aie2_config_cu(hwctx, NULL);
 	if (ret) {
 		XDNA_ERR(xdna, "Config cu failed, ret %d", ret);
 		goto out;
@@ -167,14 +168,11 @@ static int aie2_hwctx_resume_cb(struct amdxdna_hwctx *hwctx, void *arg)
 
 int aie2_hwctx_resume(struct amdxdna_client *client)
 {
-	struct amdxdna_dev *xdna = client->xdna;
-
 	/*
 	 * The resume path cannot guarantee that mailbox channel can be
 	 * regenerated. If this happen, when submit message to this
 	 * mailbox channel, error will return.
 	 */
-	drm_WARN_ON(&xdna->ddev, !mutex_is_locked(&xdna->dev_lock));
 	return amdxdna_hwctx_walk(client, NULL, aie2_hwctx_resume_cb);
 }
 
@@ -184,6 +182,8 @@ aie2_sched_notify(struct amdxdna_sched_job *job)
 	struct dma_fence *fence = job->fence;
 
 	trace_xdna_job(&job->base, job->hwctx->name, "signaled fence", job->seq);
+
+	amdxdna_pm_suspend_put(job->hwctx->client->xdna);
 	job->hwctx->priv->completed++;
 	dma_fence_signal(fence);
 
@@ -531,7 +531,7 @@ int aie2_hwctx_init(struct amdxdna_hwctx *hwctx)
 		.num_rqs = DRM_SCHED_PRIORITY_COUNT,
 		.credit_limit = HWCTX_MAX_CMDS,
 		.timeout = msecs_to_jiffies(HWCTX_MAX_TIMEOUT),
-		.name = hwctx->name,
+		.name = "amdxdna_js",
 		.dev = xdna->ddev.dev,
 	};
 	struct drm_gpu_scheduler *sched;
@@ -697,6 +697,14 @@ void aie2_hwctx_fini(struct amdxdna_hwctx *hwctx)
 	kfree(hwctx->cus);
 }
 
+static int aie2_config_cu_resp_handler(void *handle, void __iomem *data, size_t size)
+{
+	struct amdxdna_hwctx *hwctx = handle;
+
+	amdxdna_pm_suspend_put(hwctx->client->xdna);
+	return 0;
+}
+
 static int aie2_hwctx_cu_config(struct amdxdna_hwctx *hwctx, void *buf, u32 size)
 {
 	struct amdxdna_hwctx_param_config_cu *config = buf;
@@ -728,10 +736,14 @@ static int aie2_hwctx_cu_config(struct amdxdna_hwctx *hwctx, void *buf, u32 size
 	if (!hwctx->cus)
 		return -ENOMEM;
 
-	ret = aie2_config_cu(hwctx);
+	ret = amdxdna_pm_resume_get(xdna);
+	if (ret)
+		goto free_cus;
+
+	ret = aie2_config_cu(hwctx, aie2_config_cu_resp_handler);
 	if (ret) {
 		XDNA_ERR(xdna, "Config CU to firmware failed, ret %d", ret);
-		goto free_cus;
+		goto pm_suspend_put;
 	}
 
 	wmb(); /* To avoid locking in command submit when check status */
@@ -739,6 +751,8 @@ static int aie2_hwctx_cu_config(struct amdxdna_hwctx *hwctx, void *buf, u32 size
 
 	return 0;
 
+pm_suspend_put:
+	amdxdna_pm_suspend_put(xdna);
 free_cus:
 	kfree(hwctx->cus);
 	hwctx->cus = NULL;
@@ -862,11 +876,15 @@ int aie2_cmd_submit(struct amdxdna_hwctx *hwctx, struct amdxdna_sched_job *job,
 		goto free_chain;
 	}
 
+	ret = amdxdna_pm_resume_get(xdna);
+	if (ret)
+		goto cleanup_job;
+
 retry:
 	ret = drm_gem_lock_reservations(job->bos, job->bo_cnt, &acquire_ctx);
 	if (ret) {
 		XDNA_WARN(xdna, "Failed to lock BOs, ret %d", ret);
-		goto cleanup_job;
+		goto suspend_put;
 	}
 
 	for (i = 0; i < job->bo_cnt; i++) {
@@ -874,7 +892,7 @@ retry:
 		if (ret) {
 			XDNA_WARN(xdna, "Failed to reserve fences %d", ret);
 			drm_gem_unlock_reservations(job->bos, job->bo_cnt, &acquire_ctx);
-			goto cleanup_job;
+			goto suspend_put;
 		}
 	}
 
@@ -889,12 +907,12 @@ retry:
 					msecs_to_jiffies(HMM_RANGE_DEFAULT_TIMEOUT);
 			} else if (time_after(jiffies, timeout)) {
 				ret = -ETIME;
-				goto cleanup_job;
+				goto suspend_put;
 			}
 
 			ret = aie2_populate_range(abo);
 			if (ret)
-				goto cleanup_job;
+				goto suspend_put;
 			goto retry;
 		}
 	}
@@ -920,6 +938,8 @@ retry:
 
 	return 0;
 
+suspend_put:
+	amdxdna_pm_suspend_put(xdna);
 cleanup_job:
 	drm_sched_job_cleanup(&job->base);
 free_chain:
diff --git a/drivers/accel/amdxdna/aie2_message.c b/drivers/accel/amdxdna/aie2_message.c
index 9caad083543d..4660e8297ed8 100644
--- a/drivers/accel/amdxdna/aie2_message.c
+++ b/drivers/accel/amdxdna/aie2_message.c
@@ -37,7 +37,7 @@ static int aie2_send_mgmt_msg_wait(struct amdxdna_dev_hdl *ndev,
 	if (!ndev->mgmt_chann)
 		return -ENODEV;
 
-	drm_WARN_ON(&xdna->ddev, !mutex_is_locked(&xdna->dev_lock));
+	drm_WARN_ON(&xdna->ddev, xdna->rpm_on && !mutex_is_locked(&xdna->dev_lock));
 	ret = xdna_send_msg_wait(xdna, ndev->mgmt_chann, msg);
 	if (ret == -ETIME) {
 		xdna_mailbox_stop_channel(ndev->mgmt_chann);
@@ -377,15 +377,17 @@ int aie2_register_asyn_event_msg(struct amdxdna_dev_hdl *ndev, dma_addr_t addr,
 	return xdna_mailbox_send_msg(ndev->mgmt_chann, &msg, TX_TIMEOUT);
 }
 
-int aie2_config_cu(struct amdxdna_hwctx *hwctx)
+int aie2_config_cu(struct amdxdna_hwctx *hwctx,
+		   int (*notify_cb)(void *, void __iomem *, size_t))
 {
 	struct mailbox_channel *chann = hwctx->priv->mbox_chann;
 	struct amdxdna_dev *xdna = hwctx->client->xdna;
 	u32 shift = xdna->dev_info->dev_mem_buf_shift;
-	DECLARE_AIE2_MSG(config_cu, MSG_OP_CONFIG_CU);
+	struct config_cu_req req = { 0 };
+	struct xdna_mailbox_msg msg;
 	struct drm_gem_object *gobj;
 	struct amdxdna_gem_obj *abo;
-	int ret, i;
+	int i;
 
 	if (!chann)
 		return -ENODEV;
@@ -423,18 +425,12 @@ int aie2_config_cu(struct amdxdna_hwctx *hwctx)
 	}
 	req.num_cus = hwctx->cus->num_cus;
 
-	ret = xdna_send_msg_wait(xdna, chann, &msg);
-	if (ret == -ETIME)
-		aie2_destroy_context(xdna->dev_handle, hwctx);
-
-	if (resp.status == AIE2_STATUS_SUCCESS) {
-		XDNA_DBG(xdna, "Configure %d CUs, ret %d", req.num_cus, ret);
-		return 0;
-	}
-
-	XDNA_ERR(xdna, "Command opcode 0x%x failed, status 0x%x ret %d",
-		 msg.opcode, resp.status, ret);
-	return ret;
+	msg.send_data = (u8 *)&req;
+	msg.send_size = sizeof(req);
+	msg.handle = hwctx;
+	msg.opcode = MSG_OP_CONFIG_CU;
+	msg.notify_cb = notify_cb;
+	return xdna_mailbox_send_msg(chann, &msg, TX_TIMEOUT);
 }
 
 int aie2_execbuf(struct amdxdna_hwctx *hwctx, struct amdxdna_sched_job *job,
diff --git a/drivers/accel/amdxdna/aie2_pci.c b/drivers/accel/amdxdna/aie2_pci.c
index 87c425e3d2b9..8a66f276100e 100644
--- a/drivers/accel/amdxdna/aie2_pci.c
+++ b/drivers/accel/amdxdna/aie2_pci.c
@@ -25,6 +25,7 @@
 #include "amdxdna_gem.h"
 #include "amdxdna_mailbox.h"
 #include "amdxdna_pci_drv.h"
+#include "amdxdna_pm.h"
 
 static int aie2_max_col = XRS_MAX_COL;
 module_param(aie2_max_col, uint, 0600);
@@ -223,15 +224,6 @@ static int aie2_mgmt_fw_init(struct amdxdna_dev_hdl *ndev)
 		return ret;
 	}
 
-	if (!ndev->async_events)
-		return 0;
-
-	ret = aie2_error_async_events_send(ndev);
-	if (ret) {
-		XDNA_ERR(ndev->xdna, "Send async events failed");
-		return ret;
-	}
-
 	return 0;
 }
 
@@ -257,6 +249,8 @@ static int aie2_mgmt_fw_query(struct amdxdna_dev_hdl *ndev)
 		return ret;
 	}
 
+	ndev->total_col = min(aie2_max_col, ndev->metadata.cols);
+
 	return 0;
 }
 
@@ -338,6 +332,7 @@ static void aie2_hw_stop(struct amdxdna_dev *xdna)
 	ndev->mbox = NULL;
 	aie2_psp_stop(ndev->psp_hdl);
 	aie2_smu_fini(ndev);
+	aie2_error_async_events_free(ndev);
 	pci_disable_device(pdev);
 
 	ndev->dev_status = AIE2_DEV_INIT;
@@ -424,6 +419,18 @@ static int aie2_hw_start(struct amdxdna_dev *xdna)
 		goto destroy_mgmt_chann;
 	}
 
+	ret = aie2_mgmt_fw_query(ndev);
+	if (ret) {
+		XDNA_ERR(xdna, "failed to query fw, ret %d", ret);
+		goto destroy_mgmt_chann;
+	}
+
+	ret = aie2_error_async_events_alloc(ndev);
+	if (ret) {
+		XDNA_ERR(xdna, "Allocate async events failed, ret %d", ret);
+		goto destroy_mgmt_chann;
+	}
+
 	ndev->dev_status = AIE2_DEV_START;
 
 	return 0;
@@ -459,7 +466,6 @@ static int aie2_hw_resume(struct amdxdna_dev *xdna)
 	struct amdxdna_client *client;
 	int ret;
 
-	guard(mutex)(&xdna->dev_lock);
 	ret = aie2_hw_start(xdna);
 	if (ret) {
 		XDNA_ERR(xdna, "Start hardware failed, %d", ret);
@@ -565,13 +571,6 @@ static int aie2_init(struct amdxdna_dev *xdna)
 		goto release_fw;
 	}
 
-	ret = aie2_mgmt_fw_query(ndev);
-	if (ret) {
-		XDNA_ERR(xdna, "Query firmware failed, ret %d", ret);
-		goto stop_hw;
-	}
-	ndev->total_col = min(aie2_max_col, ndev->metadata.cols);
-
 	xrs_cfg.clk_list.num_levels = ndev->max_dpm_level + 1;
 	for (i = 0; i < xrs_cfg.clk_list.num_levels; i++)
 		xrs_cfg.clk_list.cu_clk_list[i] = ndev->priv->dpm_clk_tbl[i].hclk;
@@ -587,30 +586,10 @@ static int aie2_init(struct amdxdna_dev *xdna)
 		goto stop_hw;
 	}
 
-	ret = aie2_error_async_events_alloc(ndev);
-	if (ret) {
-		XDNA_ERR(xdna, "Allocate async events failed, ret %d", ret);
-		goto stop_hw;
-	}
-
-	ret = aie2_error_async_events_send(ndev);
-	if (ret) {
-		XDNA_ERR(xdna, "Send async events failed, ret %d", ret);
-		goto async_event_free;
-	}
-
-	/* Issue a command to make sure firmware handled async events */
-	ret = aie2_query_firmware_version(ndev, &ndev->xdna->fw_ver);
-	if (ret) {
-		XDNA_ERR(xdna, "Re-query firmware version failed");
-		goto async_event_free;
-	}
-
 	release_firmware(fw);
+	amdxdna_pm_init(xdna);
 	return 0;
 
-async_event_free:
-	aie2_error_async_events_free(ndev);
 stop_hw:
 	aie2_hw_stop(xdna);
 release_fw:
@@ -621,10 +600,8 @@ release_fw:
 
 static void aie2_fini(struct amdxdna_dev *xdna)
 {
-	struct amdxdna_dev_hdl *ndev = xdna->dev_handle;
-
+	amdxdna_pm_fini(xdna);
 	aie2_hw_stop(xdna);
-	aie2_error_async_events_free(ndev);
 }
 
 static int aie2_get_aie_status(struct amdxdna_client *client,
@@ -856,6 +833,10 @@ static int aie2_get_info(struct amdxdna_client *client, struct amdxdna_drm_get_i
 	if (!drm_dev_enter(&xdna->ddev, &idx))
 		return -ENODEV;
 
+	ret = amdxdna_pm_resume_get(xdna);
+	if (ret)
+		goto dev_exit;
+
 	switch (args->param) {
 	case DRM_AMDXDNA_QUERY_AIE_STATUS:
 		ret = aie2_get_aie_status(client, args);
@@ -882,8 +863,11 @@ static int aie2_get_info(struct amdxdna_client *client, struct amdxdna_drm_get_i
 		XDNA_ERR(xdna, "Not supported request parameter %u", args->param);
 		ret = -EOPNOTSUPP;
 	}
+
+	amdxdna_pm_suspend_put(xdna);
 	XDNA_DBG(xdna, "Got param %d", args->param);
 
+dev_exit:
 	drm_dev_exit(idx);
 	return ret;
 }
@@ -898,6 +882,12 @@ static int aie2_query_ctx_status_array(struct amdxdna_client *client,
 
 	drm_WARN_ON(&xdna->ddev, !mutex_is_locked(&xdna->dev_lock));
 
+	if (args->element_size > SZ_4K || args->num_element > SZ_1K) {
+		XDNA_DBG(xdna, "Invalid element size %d or number of element %d",
+			 args->element_size, args->num_element);
+		return -EINVAL;
+	}
+
 	array_args.element_size = min(args->element_size,
 				      sizeof(struct amdxdna_drm_hwctx_entry));
 	array_args.buffer = args->buffer;
@@ -926,6 +916,10 @@ static int aie2_get_array(struct amdxdna_client *client,
 	if (!drm_dev_enter(&xdna->ddev, &idx))
 		return -ENODEV;
 
+	ret = amdxdna_pm_resume_get(xdna);
+	if (ret)
+		goto dev_exit;
+
 	switch (args->param) {
 	case DRM_AMDXDNA_HW_CONTEXT_ALL:
 		ret = aie2_query_ctx_status_array(client, args);
@@ -934,8 +928,11 @@ static int aie2_get_array(struct amdxdna_client *client,
 		XDNA_ERR(xdna, "Not supported request parameter %u", args->param);
 		ret = -EOPNOTSUPP;
 	}
+
+	amdxdna_pm_suspend_put(xdna);
 	XDNA_DBG(xdna, "Got param %d", args->param);
 
+dev_exit:
 	drm_dev_exit(idx);
 	return ret;
 }
@@ -974,6 +971,10 @@ static int aie2_set_state(struct amdxdna_client *client,
 	if (!drm_dev_enter(&xdna->ddev, &idx))
 		return -ENODEV;
 
+	ret = amdxdna_pm_resume_get(xdna);
+	if (ret)
+		goto dev_exit;
+
 	switch (args->param) {
 	case DRM_AMDXDNA_SET_POWER_MODE:
 		ret = aie2_set_power_mode(client, args);
@@ -984,6 +985,8 @@ static int aie2_set_state(struct amdxdna_client *client,
 		break;
 	}
 
+	amdxdna_pm_suspend_put(xdna);
+dev_exit:
 	drm_dev_exit(idx);
 	return ret;
 }
diff --git a/drivers/accel/amdxdna/aie2_pci.h b/drivers/accel/amdxdna/aie2_pci.h
index 91a8e948f82a..289a23ecd5f1 100644
--- a/drivers/accel/amdxdna/aie2_pci.h
+++ b/drivers/accel/amdxdna/aie2_pci.h
@@ -272,7 +272,8 @@ int aie2_map_host_buf(struct amdxdna_dev_hdl *ndev, u32 context_id, u64 addr, u6
 int aie2_query_status(struct amdxdna_dev_hdl *ndev, char __user *buf, u32 size, u32 *cols_filled);
 int aie2_register_asyn_event_msg(struct amdxdna_dev_hdl *ndev, dma_addr_t addr, u32 size,
 				 void *handle, int (*cb)(void*, void __iomem *, size_t));
-int aie2_config_cu(struct amdxdna_hwctx *hwctx);
+int aie2_config_cu(struct amdxdna_hwctx *hwctx,
+		   int (*notify_cb)(void *, void __iomem *, size_t));
 int aie2_execbuf(struct amdxdna_hwctx *hwctx, struct amdxdna_sched_job *job,
 		 int (*notify_cb)(void *, void __iomem *, size_t));
 int aie2_cmdlist_single_execbuf(struct amdxdna_hwctx *hwctx,
diff --git a/drivers/accel/amdxdna/aie2_smu.c b/drivers/accel/amdxdna/aie2_smu.c
index d303701b0ded..7f292a615ed8 100644
--- a/drivers/accel/amdxdna/aie2_smu.c
+++ b/drivers/accel/amdxdna/aie2_smu.c
@@ -11,6 +11,7 @@
 
 #include "aie2_pci.h"
 #include "amdxdna_pci_drv.h"
+#include "amdxdna_pm.h"
 
 #define SMU_RESULT_OK		1
 
@@ -59,12 +60,16 @@ int npu1_set_dpm(struct amdxdna_dev_hdl *ndev, u32 dpm_level)
 	u32 freq;
 	int ret;
 
+	ret = amdxdna_pm_resume_get(ndev->xdna);
+	if (ret)
+		return ret;
+
 	ret = aie2_smu_exec(ndev, AIE2_SMU_SET_MPNPUCLK_FREQ,
 			    ndev->priv->dpm_clk_tbl[dpm_level].npuclk, &freq);
 	if (ret) {
 		XDNA_ERR(ndev->xdna, "Set npu clock to %d failed, ret %d\n",
 			 ndev->priv->dpm_clk_tbl[dpm_level].npuclk, ret);
-		return ret;
+		goto suspend_put;
 	}
 	ndev->npuclk_freq = freq;
 
@@ -73,8 +78,10 @@ int npu1_set_dpm(struct amdxdna_dev_hdl *ndev, u32 dpm_level)
 	if (ret) {
 		XDNA_ERR(ndev->xdna, "Set h clock to %d failed, ret %d\n",
 			 ndev->priv->dpm_clk_tbl[dpm_level].hclk, ret);
-		return ret;
+		goto suspend_put;
 	}
+
+	amdxdna_pm_suspend_put(ndev->xdna);
 	ndev->hclk_freq = freq;
 	ndev->dpm_level = dpm_level;
 
@@ -82,26 +89,35 @@ int npu1_set_dpm(struct amdxdna_dev_hdl *ndev, u32 dpm_level)
 		 ndev->npuclk_freq, ndev->hclk_freq);
 
 	return 0;
+
+suspend_put:
+	amdxdna_pm_suspend_put(ndev->xdna);
+	return ret;
 }
 
 int npu4_set_dpm(struct amdxdna_dev_hdl *ndev, u32 dpm_level)
 {
 	int ret;
 
+	ret = amdxdna_pm_resume_get(ndev->xdna);
+	if (ret)
+		return ret;
+
 	ret = aie2_smu_exec(ndev, AIE2_SMU_SET_HARD_DPMLEVEL, dpm_level, NULL);
 	if (ret) {
 		XDNA_ERR(ndev->xdna, "Set hard dpm level %d failed, ret %d ",
 			 dpm_level, ret);
-		return ret;
+		goto suspend_put;
 	}
 
 	ret = aie2_smu_exec(ndev, AIE2_SMU_SET_SOFT_DPMLEVEL, dpm_level, NULL);
 	if (ret) {
 		XDNA_ERR(ndev->xdna, "Set soft dpm level %d failed, ret %d",
 			 dpm_level, ret);
-		return ret;
+		goto suspend_put;
 	}
 
+	amdxdna_pm_suspend_put(ndev->xdna);
 	ndev->npuclk_freq = ndev->priv->dpm_clk_tbl[dpm_level].npuclk;
 	ndev->hclk_freq = ndev->priv->dpm_clk_tbl[dpm_level].hclk;
 	ndev->dpm_level = dpm_level;
@@ -110,6 +126,10 @@ int npu4_set_dpm(struct amdxdna_dev_hdl *ndev, u32 dpm_level)
 		 ndev->npuclk_freq, ndev->hclk_freq);
 
 	return 0;
+
+suspend_put:
+	amdxdna_pm_suspend_put(ndev->xdna);
+	return ret;
 }
 
 int aie2_smu_init(struct amdxdna_dev_hdl *ndev)
diff --git a/drivers/accel/amdxdna/amdxdna_ctx.c b/drivers/accel/amdxdna/amdxdna_ctx.c
index 4bfe4ef20550..868ca369e0a0 100644
--- a/drivers/accel/amdxdna/amdxdna_ctx.c
+++ b/drivers/accel/amdxdna/amdxdna_ctx.c
@@ -161,19 +161,14 @@ int amdxdna_drm_create_hwctx_ioctl(struct drm_device *dev, void *data, struct dr
 	if (args->ext || args->ext_flags)
 		return -EINVAL;
 
-	if (!drm_dev_enter(dev, &idx))
-		return -ENODEV;
-
 	hwctx = kzalloc(sizeof(*hwctx), GFP_KERNEL);
-	if (!hwctx) {
-		ret = -ENOMEM;
-		goto exit;
-	}
+	if (!hwctx)
+		return -ENOMEM;
 
 	if (copy_from_user(&hwctx->qos, u64_to_user_ptr(args->qos_p), sizeof(hwctx->qos))) {
 		XDNA_ERR(xdna, "Access QoS info failed");
-		ret = -EFAULT;
-		goto free_hwctx;
+		kfree(hwctx);
+		return -EFAULT;
 	}
 
 	hwctx->client = client;
@@ -181,30 +176,36 @@ int amdxdna_drm_create_hwctx_ioctl(struct drm_device *dev, void *data, struct dr
 	hwctx->num_tiles = args->num_tiles;
 	hwctx->mem_size = args->mem_size;
 	hwctx->max_opc = args->max_opc;
-	ret = xa_alloc_cyclic(&client->hwctx_xa, &hwctx->id, hwctx,
-			      XA_LIMIT(AMDXDNA_INVALID_CTX_HANDLE + 1, MAX_HWCTX_ID),
-			      &client->next_hwctxid, GFP_KERNEL);
-	if (ret < 0) {
-		XDNA_ERR(xdna, "Allocate hwctx ID failed, ret %d", ret);
+
+	guard(mutex)(&xdna->dev_lock);
+
+	if (!drm_dev_enter(dev, &idx)) {
+		ret = -ENODEV;
 		goto free_hwctx;
 	}
 
-	hwctx->name = kasprintf(GFP_KERNEL, "hwctx.%d.%d", client->pid, hwctx->id);
+	ret = xdna->dev_info->ops->hwctx_init(hwctx);
+	if (ret) {
+		XDNA_ERR(xdna, "Init hwctx failed, ret %d", ret);
+		goto dev_exit;
+	}
+
+	hwctx->name = kasprintf(GFP_KERNEL, "hwctx.%d.%d", client->pid, hwctx->fw_ctx_id);
 	if (!hwctx->name) {
 		ret = -ENOMEM;
-		goto rm_id;
+		goto fini_hwctx;
 	}
 
-	mutex_lock(&xdna->dev_lock);
-	ret = xdna->dev_info->ops->hwctx_init(hwctx);
-	if (ret) {
-		mutex_unlock(&xdna->dev_lock);
-		XDNA_ERR(xdna, "Init hwctx failed, ret %d", ret);
+	ret = xa_alloc_cyclic(&client->hwctx_xa, &hwctx->id, hwctx,
+			      XA_LIMIT(AMDXDNA_INVALID_CTX_HANDLE + 1, MAX_HWCTX_ID),
+			      &client->next_hwctxid, GFP_KERNEL);
+	if (ret < 0) {
+		XDNA_ERR(xdna, "Allocate hwctx ID failed, ret %d", ret);
 		goto free_name;
 	}
+
 	args->handle = hwctx->id;
 	args->syncobj_handle = hwctx->syncobj_hdl;
-	mutex_unlock(&xdna->dev_lock);
 
 	atomic64_set(&hwctx->job_submit_cnt, 0);
 	atomic64_set(&hwctx->job_free_cnt, 0);
@@ -214,12 +215,12 @@ int amdxdna_drm_create_hwctx_ioctl(struct drm_device *dev, void *data, struct dr
 
 free_name:
 	kfree(hwctx->name);
-rm_id:
-	xa_erase(&client->hwctx_xa, hwctx->id);
+fini_hwctx:
+	xdna->dev_info->ops->hwctx_fini(hwctx);
+dev_exit:
+	drm_dev_exit(idx);
 free_hwctx:
 	kfree(hwctx);
-exit:
-	drm_dev_exit(idx);
 	return ret;
 }
 
@@ -431,11 +432,6 @@ int amdxdna_cmd_submit(struct amdxdna_client *client,
 		goto unlock_srcu;
 	}
 
-	if (hwctx->status != HWCTX_STAT_READY) {
-		XDNA_ERR(xdna, "HW Context is not ready");
-		ret = -EINVAL;
-		goto unlock_srcu;
-	}
 
 	job->hwctx = hwctx;
 	job->mm = current->mm;
diff --git a/drivers/accel/amdxdna/amdxdna_gem.c b/drivers/accel/amdxdna/amdxdna_gem.c
index d407a36eb412..7f91863c3f24 100644
--- a/drivers/accel/amdxdna/amdxdna_gem.c
+++ b/drivers/accel/amdxdna/amdxdna_gem.c
@@ -392,35 +392,33 @@ static const struct dma_buf_ops amdxdna_dmabuf_ops = {
 	.vunmap = drm_gem_dmabuf_vunmap,
 };
 
-static int amdxdna_gem_obj_vmap(struct drm_gem_object *obj, struct iosys_map *map)
+static int amdxdna_gem_obj_vmap(struct amdxdna_gem_obj *abo, void **vaddr)
 {
-	struct amdxdna_gem_obj *abo = to_xdna_obj(obj);
-
-	iosys_map_clear(map);
-
-	dma_resv_assert_held(obj->resv);
+	struct iosys_map map = IOSYS_MAP_INIT_VADDR(NULL);
+	int ret;
 
 	if (is_import_bo(abo))
-		dma_buf_vmap(abo->dma_buf, map);
+		ret = dma_buf_vmap_unlocked(abo->dma_buf, &map);
 	else
-		drm_gem_shmem_object_vmap(obj, map);
-
-	if (!map->vaddr)
-		return -ENOMEM;
+		ret = drm_gem_vmap(to_gobj(abo), &map);
 
-	return 0;
+	*vaddr = map.vaddr;
+	return ret;
 }
 
-static void amdxdna_gem_obj_vunmap(struct drm_gem_object *obj, struct iosys_map *map)
+static void amdxdna_gem_obj_vunmap(struct amdxdna_gem_obj *abo)
 {
-	struct amdxdna_gem_obj *abo = to_xdna_obj(obj);
+	struct iosys_map map;
+
+	if (!abo->mem.kva)
+		return;
 
-	dma_resv_assert_held(obj->resv);
+	iosys_map_set_vaddr(&map, abo->mem.kva);
 
 	if (is_import_bo(abo))
-		dma_buf_vunmap(abo->dma_buf, map);
+		dma_buf_vunmap_unlocked(abo->dma_buf, &map);
 	else
-		drm_gem_shmem_object_vunmap(obj, map);
+		drm_gem_vunmap(to_gobj(abo), &map);
 }
 
 static struct dma_buf *amdxdna_gem_prime_export(struct drm_gem_object *gobj, int flags)
@@ -455,7 +453,6 @@ static void amdxdna_gem_obj_free(struct drm_gem_object *gobj)
 {
 	struct amdxdna_dev *xdna = to_xdna_dev(gobj->dev);
 	struct amdxdna_gem_obj *abo = to_xdna_obj(gobj);
-	struct iosys_map map = IOSYS_MAP_INIT_VADDR(abo->mem.kva);
 
 	XDNA_DBG(xdna, "BO type %d xdna_addr 0x%llx", abo->type, abo->mem.dev_addr);
 
@@ -468,7 +465,7 @@ static void amdxdna_gem_obj_free(struct drm_gem_object *gobj)
 	if (abo->type == AMDXDNA_BO_DEV_HEAP)
 		drm_mm_takedown(&abo->mm);
 
-	drm_gem_vunmap(gobj, &map);
+	amdxdna_gem_obj_vunmap(abo);
 	mutex_destroy(&abo->lock);
 
 	if (is_import_bo(abo)) {
@@ -489,8 +486,8 @@ static const struct drm_gem_object_funcs amdxdna_gem_shmem_funcs = {
 	.pin = drm_gem_shmem_object_pin,
 	.unpin = drm_gem_shmem_object_unpin,
 	.get_sg_table = drm_gem_shmem_object_get_sg_table,
-	.vmap = amdxdna_gem_obj_vmap,
-	.vunmap = amdxdna_gem_obj_vunmap,
+	.vmap = drm_gem_shmem_object_vmap,
+	.vunmap = drm_gem_shmem_object_vunmap,
 	.mmap = amdxdna_gem_obj_mmap,
 	.vm_ops = &drm_gem_shmem_vm_ops,
 	.export = amdxdna_gem_prime_export,
@@ -663,7 +660,6 @@ amdxdna_drm_create_dev_heap(struct drm_device *dev,
 			    struct drm_file *filp)
 {
 	struct amdxdna_client *client = filp->driver_priv;
-	struct iosys_map map = IOSYS_MAP_INIT_VADDR(NULL);
 	struct amdxdna_dev *xdna = to_xdna_dev(dev);
 	struct amdxdna_gem_obj *abo;
 	int ret;
@@ -692,12 +688,11 @@ amdxdna_drm_create_dev_heap(struct drm_device *dev,
 	abo->mem.dev_addr = client->xdna->dev_info->dev_mem_base;
 	drm_mm_init(&abo->mm, abo->mem.dev_addr, abo->mem.size);
 
-	ret = drm_gem_vmap(to_gobj(abo), &map);
+	ret = amdxdna_gem_obj_vmap(abo, &abo->mem.kva);
 	if (ret) {
 		XDNA_ERR(xdna, "Vmap heap bo failed, ret %d", ret);
 		goto release_obj;
 	}
-	abo->mem.kva = map.vaddr;
 
 	client->dev_heap = abo;
 	drm_gem_object_get(to_gobj(abo));
@@ -748,7 +743,6 @@ amdxdna_drm_create_cmd_bo(struct drm_device *dev,
 			  struct amdxdna_drm_create_bo *args,
 			  struct drm_file *filp)
 {
-	struct iosys_map map = IOSYS_MAP_INIT_VADDR(NULL);
 	struct amdxdna_dev *xdna = to_xdna_dev(dev);
 	struct amdxdna_gem_obj *abo;
 	int ret;
@@ -770,12 +764,11 @@ amdxdna_drm_create_cmd_bo(struct drm_device *dev,
 	abo->type = AMDXDNA_BO_CMD;
 	abo->client = filp->driver_priv;
 
-	ret = drm_gem_vmap(to_gobj(abo), &map);
+	ret = amdxdna_gem_obj_vmap(abo, &abo->mem.kva);
 	if (ret) {
 		XDNA_ERR(xdna, "Vmap cmd bo failed, ret %d", ret);
 		goto release_obj;
 	}
-	abo->mem.kva = map.vaddr;
 
 	return abo;
 
diff --git a/drivers/accel/amdxdna/amdxdna_mailbox.c b/drivers/accel/amdxdna/amdxdna_mailbox.c
index da1ac89bb78f..24258dcc18eb 100644
--- a/drivers/accel/amdxdna/amdxdna_mailbox.c
+++ b/drivers/accel/amdxdna/amdxdna_mailbox.c
@@ -194,7 +194,8 @@ static void mailbox_release_msg(struct mailbox_channel *mb_chann,
 {
 	MB_DBG(mb_chann, "msg_id 0x%x msg opcode 0x%x",
 	       mb_msg->pkg.header.id, mb_msg->pkg.header.opcode);
-	mb_msg->notify_cb(mb_msg->handle, NULL, 0);
+	if (mb_msg->notify_cb)
+		mb_msg->notify_cb(mb_msg->handle, NULL, 0);
 	kfree(mb_msg);
 }
 
@@ -248,7 +249,7 @@ mailbox_get_resp(struct mailbox_channel *mb_chann, struct xdna_msg_header *heade
 {
 	struct mailbox_msg *mb_msg;
 	int msg_id;
-	int ret;
+	int ret = 0;
 
 	msg_id = header->id;
 	if (!mailbox_validate_msgid(msg_id)) {
@@ -265,9 +266,11 @@ mailbox_get_resp(struct mailbox_channel *mb_chann, struct xdna_msg_header *heade
 
 	MB_DBG(mb_chann, "opcode 0x%x size %d id 0x%x",
 	       header->opcode, header->total_size, header->id);
-	ret = mb_msg->notify_cb(mb_msg->handle, data, header->total_size);
-	if (unlikely(ret))
-		MB_ERR(mb_chann, "Message callback ret %d", ret);
+	if (mb_msg->notify_cb) {
+		ret = mb_msg->notify_cb(mb_msg->handle, data, header->total_size);
+		if (unlikely(ret))
+			MB_ERR(mb_chann, "Message callback ret %d", ret);
+	}
 
 	kfree(mb_msg);
 	return ret;
diff --git a/drivers/accel/amdxdna/amdxdna_pci_drv.c b/drivers/accel/amdxdna/amdxdna_pci_drv.c
index 569cd703729d..aa04452310e5 100644
--- a/drivers/accel/amdxdna/amdxdna_pci_drv.c
+++ b/drivers/accel/amdxdna/amdxdna_pci_drv.c
@@ -13,13 +13,11 @@
 #include <drm/gpu_scheduler.h>
 #include <linux/iommu.h>
 #include <linux/pci.h>
-#include <linux/pm_runtime.h>
 
 #include "amdxdna_ctx.h"
 #include "amdxdna_gem.h"
 #include "amdxdna_pci_drv.h"
-
-#define AMDXDNA_AUTOSUSPEND_DELAY	5000 /* milliseconds */
+#include "amdxdna_pm.h"
 
 MODULE_FIRMWARE("amdnpu/1502_00/npu.sbin");
 MODULE_FIRMWARE("amdnpu/17f0_10/npu.sbin");
@@ -61,17 +59,9 @@ static int amdxdna_drm_open(struct drm_device *ddev, struct drm_file *filp)
 	struct amdxdna_client *client;
 	int ret;
 
-	ret = pm_runtime_resume_and_get(ddev->dev);
-	if (ret) {
-		XDNA_ERR(xdna, "Failed to get rpm, ret %d", ret);
-		return ret;
-	}
-
 	client = kzalloc(sizeof(*client), GFP_KERNEL);
-	if (!client) {
-		ret = -ENOMEM;
-		goto put_rpm;
-	}
+	if (!client)
+		return -ENOMEM;
 
 	client->pid = pid_nr(rcu_access_pointer(filp->pid));
 	client->xdna = xdna;
@@ -106,9 +96,6 @@ unbind_sva:
 	iommu_sva_unbind_device(client->sva);
 failed:
 	kfree(client);
-put_rpm:
-	pm_runtime_mark_last_busy(ddev->dev);
-	pm_runtime_put_autosuspend(ddev->dev);
 
 	return ret;
 }
@@ -130,8 +117,6 @@ static void amdxdna_drm_close(struct drm_device *ddev, struct drm_file *filp)
 
 	XDNA_DBG(xdna, "pid %d closed", client->pid);
 	kfree(client);
-	pm_runtime_mark_last_busy(ddev->dev);
-	pm_runtime_put_autosuspend(ddev->dev);
 }
 
 static int amdxdna_flush(struct file *f, fl_owner_t id)
@@ -310,19 +295,12 @@ static int amdxdna_probe(struct pci_dev *pdev, const struct pci_device_id *id)
 		goto failed_dev_fini;
 	}
 
-	pm_runtime_set_autosuspend_delay(dev, AMDXDNA_AUTOSUSPEND_DELAY);
-	pm_runtime_use_autosuspend(dev);
-	pm_runtime_allow(dev);
-
 	ret = drm_dev_register(&xdna->ddev, 0);
 	if (ret) {
 		XDNA_ERR(xdna, "DRM register failed, ret %d", ret);
-		pm_runtime_forbid(dev);
 		goto failed_sysfs_fini;
 	}
 
-	pm_runtime_mark_last_busy(dev);
-	pm_runtime_put_autosuspend(dev);
 	return 0;
 
 failed_sysfs_fini:
@@ -339,14 +317,10 @@ destroy_notifier_wq:
 static void amdxdna_remove(struct pci_dev *pdev)
 {
 	struct amdxdna_dev *xdna = pci_get_drvdata(pdev);
-	struct device *dev = &pdev->dev;
 	struct amdxdna_client *client;
 
 	destroy_workqueue(xdna->notifier_wq);
 
-	pm_runtime_get_noresume(dev);
-	pm_runtime_forbid(dev);
-
 	drm_dev_unplug(&xdna->ddev);
 	amdxdna_sysfs_fini(xdna);
 
@@ -365,29 +339,9 @@ static void amdxdna_remove(struct pci_dev *pdev)
 	mutex_unlock(&xdna->dev_lock);
 }
 
-static int amdxdna_pmops_suspend(struct device *dev)
-{
-	struct amdxdna_dev *xdna = pci_get_drvdata(to_pci_dev(dev));
-
-	if (!xdna->dev_info->ops->suspend)
-		return -EOPNOTSUPP;
-
-	return xdna->dev_info->ops->suspend(xdna);
-}
-
-static int amdxdna_pmops_resume(struct device *dev)
-{
-	struct amdxdna_dev *xdna = pci_get_drvdata(to_pci_dev(dev));
-
-	if (!xdna->dev_info->ops->resume)
-		return -EOPNOTSUPP;
-
-	return xdna->dev_info->ops->resume(xdna);
-}
-
 static const struct dev_pm_ops amdxdna_pm_ops = {
-	SYSTEM_SLEEP_PM_OPS(amdxdna_pmops_suspend, amdxdna_pmops_resume)
-	RUNTIME_PM_OPS(amdxdna_pmops_suspend, amdxdna_pmops_resume, NULL)
+	SYSTEM_SLEEP_PM_OPS(amdxdna_pm_suspend, amdxdna_pm_resume)
+	RUNTIME_PM_OPS(amdxdna_pm_suspend, amdxdna_pm_resume, NULL)
 };
 
 static struct pci_driver amdxdna_pci_driver = {
diff --git a/drivers/accel/amdxdna/amdxdna_pci_drv.h b/drivers/accel/amdxdna/amdxdna_pci_drv.h
index 72d6696d49da..626beebf730e 100644
--- a/drivers/accel/amdxdna/amdxdna_pci_drv.h
+++ b/drivers/accel/amdxdna/amdxdna_pci_drv.h
@@ -6,6 +6,7 @@
 #ifndef _AMDXDNA_PCI_DRV_H_
 #define _AMDXDNA_PCI_DRV_H_
 
+#include <drm/drm_print.h>
 #include <linux/workqueue.h>
 #include <linux/xarray.h>
 
@@ -99,6 +100,7 @@ struct amdxdna_dev {
 	struct amdxdna_fw_ver		fw_ver;
 	struct rw_semaphore		notifier_lock; /* for mmu notifier*/
 	struct workqueue_struct		*notifier_wq;
+	bool				rpm_on;
 };
 
 /*
diff --git a/drivers/accel/amdxdna/amdxdna_pm.c b/drivers/accel/amdxdna/amdxdna_pm.c
new file mode 100644
index 000000000000..fa38e65d617c
--- /dev/null
+++ b/drivers/accel/amdxdna/amdxdna_pm.c
@@ -0,0 +1,94 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright (C) 2025, Advanced Micro Devices, Inc.
+ */
+
+#include <drm/amdxdna_accel.h>
+#include <drm/drm_drv.h>
+#include <linux/pm_runtime.h>
+
+#include "amdxdna_pm.h"
+
+#define AMDXDNA_AUTOSUSPEND_DELAY	5000 /* milliseconds */
+
+int amdxdna_pm_suspend(struct device *dev)
+{
+	struct amdxdna_dev *xdna = to_xdna_dev(dev_get_drvdata(dev));
+	int ret = -EOPNOTSUPP;
+	bool rpm;
+
+	if (xdna->dev_info->ops->suspend) {
+		rpm = xdna->rpm_on;
+		xdna->rpm_on = false;
+		ret = xdna->dev_info->ops->suspend(xdna);
+		xdna->rpm_on = rpm;
+	}
+
+	XDNA_DBG(xdna, "Suspend done ret %d", ret);
+	return ret;
+}
+
+int amdxdna_pm_resume(struct device *dev)
+{
+	struct amdxdna_dev *xdna = to_xdna_dev(dev_get_drvdata(dev));
+	int ret = -EOPNOTSUPP;
+	bool rpm;
+
+	if (xdna->dev_info->ops->resume) {
+		rpm = xdna->rpm_on;
+		xdna->rpm_on = false;
+		ret = xdna->dev_info->ops->resume(xdna);
+		xdna->rpm_on = rpm;
+	}
+
+	XDNA_DBG(xdna, "Resume done ret %d", ret);
+	return ret;
+}
+
+int amdxdna_pm_resume_get(struct amdxdna_dev *xdna)
+{
+	struct device *dev = xdna->ddev.dev;
+	int ret;
+
+	if (!xdna->rpm_on)
+		return 0;
+
+	ret = pm_runtime_resume_and_get(dev);
+	if (ret) {
+		XDNA_ERR(xdna, "Resume failed: %d", ret);
+		pm_runtime_set_suspended(dev);
+	}
+
+	return ret;
+}
+
+void amdxdna_pm_suspend_put(struct amdxdna_dev *xdna)
+{
+	struct device *dev = xdna->ddev.dev;
+
+	if (!xdna->rpm_on)
+		return;
+
+	pm_runtime_put_autosuspend(dev);
+}
+
+void amdxdna_pm_init(struct amdxdna_dev *xdna)
+{
+	struct device *dev = xdna->ddev.dev;
+
+	pm_runtime_set_active(dev);
+	pm_runtime_set_autosuspend_delay(dev, AMDXDNA_AUTOSUSPEND_DELAY);
+	pm_runtime_use_autosuspend(dev);
+	pm_runtime_allow(dev);
+	pm_runtime_put_autosuspend(dev);
+	xdna->rpm_on = true;
+}
+
+void amdxdna_pm_fini(struct amdxdna_dev *xdna)
+{
+	struct device *dev = xdna->ddev.dev;
+
+	xdna->rpm_on = false;
+	pm_runtime_get_noresume(dev);
+	pm_runtime_forbid(dev);
+}
diff --git a/drivers/accel/amdxdna/amdxdna_pm.h b/drivers/accel/amdxdna/amdxdna_pm.h
new file mode 100644
index 000000000000..77b2d6e45570
--- /dev/null
+++ b/drivers/accel/amdxdna/amdxdna_pm.h
@@ -0,0 +1,18 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/*
+ * Copyright (C) 2025, Advanced Micro Devices, Inc.
+ */
+
+#ifndef _AMDXDNA_PM_H_
+#define _AMDXDNA_PM_H_
+
+#include "amdxdna_pci_drv.h"
+
+int amdxdna_pm_suspend(struct device *dev);
+int amdxdna_pm_resume(struct device  *dev);
+int amdxdna_pm_resume_get(struct amdxdna_dev *xdna);
+void amdxdna_pm_suspend_put(struct amdxdna_dev *xdna);
+void amdxdna_pm_init(struct amdxdna_dev *xdna);
+void amdxdna_pm_fini(struct amdxdna_dev *xdna);
+
+#endif /* _AMDXDNA_PM_H_ */
diff --git a/drivers/accel/ivpu/ivpu_debugfs.c b/drivers/accel/ivpu/ivpu_debugfs.c
index cd24ccd20ba6..3bd85ee6c26b 100644
--- a/drivers/accel/ivpu/ivpu_debugfs.c
+++ b/drivers/accel/ivpu/ivpu_debugfs.c
@@ -398,35 +398,25 @@ static int dct_active_set(void *data, u64 active_percent)
 
 DEFINE_DEBUGFS_ATTRIBUTE(ivpu_dct_fops, dct_active_get, dct_active_set, "%llu\n");
 
+static void print_priority_band(struct seq_file *s, struct ivpu_hw_info *hw,
+				int band, const char *name)
+{
+	seq_printf(s, "%-9s: grace_period %9u process_grace_period %9u process_quantum %9u\n",
+		   name,
+		   hw->hws.grace_period[band],
+		   hw->hws.process_grace_period[band],
+		   hw->hws.process_quantum[band]);
+}
+
 static int priority_bands_show(struct seq_file *s, void *v)
 {
 	struct ivpu_device *vdev = s->private;
 	struct ivpu_hw_info *hw = vdev->hw;
 
-	for (int band = VPU_JOB_SCHEDULING_PRIORITY_BAND_IDLE;
-	     band < VPU_JOB_SCHEDULING_PRIORITY_BAND_COUNT; band++) {
-		switch (band) {
-		case VPU_JOB_SCHEDULING_PRIORITY_BAND_IDLE:
-			seq_puts(s, "Idle:     ");
-			break;
-
-		case VPU_JOB_SCHEDULING_PRIORITY_BAND_NORMAL:
-			seq_puts(s, "Normal:   ");
-			break;
-
-		case VPU_JOB_SCHEDULING_PRIORITY_BAND_FOCUS:
-			seq_puts(s, "Focus:    ");
-			break;
-
-		case VPU_JOB_SCHEDULING_PRIORITY_BAND_REALTIME:
-			seq_puts(s, "Realtime: ");
-			break;
-		}
-
-		seq_printf(s, "grace_period %9u process_grace_period %9u process_quantum %9u\n",
-			   hw->hws.grace_period[band], hw->hws.process_grace_period[band],
-			   hw->hws.process_quantum[band]);
-	}
+	print_priority_band(s, hw, VPU_JOB_SCHEDULING_PRIORITY_BAND_IDLE, "Idle");
+	print_priority_band(s, hw, VPU_JOB_SCHEDULING_PRIORITY_BAND_NORMAL, "Normal");
+	print_priority_band(s, hw, VPU_JOB_SCHEDULING_PRIORITY_BAND_FOCUS, "Focus");
+	print_priority_band(s, hw, VPU_JOB_SCHEDULING_PRIORITY_BAND_REALTIME, "Realtime");
 
 	return 0;
 }
diff --git a/drivers/accel/ivpu/ivpu_drv.c b/drivers/accel/ivpu/ivpu_drv.c
index 3289751b4757..1792d0bbec71 100644
--- a/drivers/accel/ivpu/ivpu_drv.c
+++ b/drivers/accel/ivpu/ivpu_drv.c
@@ -200,6 +200,9 @@ static int ivpu_get_param_ioctl(struct drm_device *dev, void *data, struct drm_f
 	case DRM_IVPU_PARAM_CAPABILITIES:
 		args->value = ivpu_is_capable(vdev, args->index);
 		break;
+	case DRM_IVPU_PARAM_PREEMPT_BUFFER_SIZE:
+		args->value = ivpu_fw_preempt_buf_size(vdev);
+		break;
 	default:
 		ret = -EINVAL;
 		break;
@@ -377,8 +380,7 @@ int ivpu_boot(struct ivpu_device *vdev)
 	drm_WARN_ON(&vdev->drm, atomic_read(&vdev->job_timeout_counter));
 	drm_WARN_ON(&vdev->drm, !xa_empty(&vdev->submitted_jobs_xa));
 
-	/* Update boot params located at first 4KB of FW memory */
-	ivpu_fw_boot_params_setup(vdev, ivpu_bo_vaddr(vdev->fw->mem));
+	ivpu_fw_boot_params_setup(vdev, ivpu_bo_vaddr(vdev->fw->mem_bp));
 
 	ret = ivpu_hw_boot_fw(vdev);
 	if (ret) {
diff --git a/drivers/accel/ivpu/ivpu_fw.c b/drivers/accel/ivpu/ivpu_fw.c
index 9db741695401..be1290be77fd 100644
--- a/drivers/accel/ivpu/ivpu_fw.c
+++ b/drivers/accel/ivpu/ivpu_fw.c
@@ -17,15 +17,10 @@
 #include "ivpu_ipc.h"
 #include "ivpu_pm.h"
 
-#define FW_GLOBAL_MEM_START	(2ull * SZ_1G)
-#define FW_GLOBAL_MEM_END	(3ull * SZ_1G)
-#define FW_SHARED_MEM_SIZE	SZ_256M /* Must be aligned to FW_SHARED_MEM_ALIGNMENT */
-#define FW_SHARED_MEM_ALIGNMENT	SZ_128K /* VPU MTRR limitation */
-#define FW_RUNTIME_MAX_SIZE	SZ_512M
 #define FW_SHAVE_NN_MAX_SIZE	SZ_2M
-#define FW_RUNTIME_MIN_ADDR	(FW_GLOBAL_MEM_START)
-#define FW_RUNTIME_MAX_ADDR	(FW_GLOBAL_MEM_END - FW_SHARED_MEM_SIZE)
 #define FW_FILE_IMAGE_OFFSET	(VPU_FW_HEADER_SIZE + FW_VERSION_HEADER_SIZE)
+#define FW_PREEMPT_BUF_MIN_SIZE SZ_4K
+#define FW_PREEMPT_BUF_MAX_SIZE SZ_32M
 
 #define WATCHDOG_MSS_REDIRECT	32
 #define WATCHDOG_NCE_REDIRECT	33
@@ -131,9 +126,14 @@ ivpu_fw_check_api_ver_lt(struct ivpu_device *vdev, const struct vpu_firmware_hea
 	return false;
 }
 
-static bool is_within_range(u64 addr, size_t size, u64 range_start, size_t range_size)
+bool ivpu_is_within_range(u64 addr, size_t size, struct ivpu_addr_range *range)
 {
-	if (addr < range_start || addr + size > range_start + range_size)
+	u64 addr_end;
+
+	if (!range || check_add_overflow(addr, size, &addr_end))
+		return false;
+
+	if (addr < range->start || addr_end > range->end)
 		return false;
 
 	return true;
@@ -151,11 +151,56 @@ ivpu_fw_sched_mode_select(struct ivpu_device *vdev, const struct vpu_firmware_he
 	return VPU_SCHEDULING_MODE_HW;
 }
 
+static void
+ivpu_preemption_config_parse(struct ivpu_device *vdev, const struct vpu_firmware_header *fw_hdr)
+{
+	struct ivpu_fw_info *fw = vdev->fw;
+	u32 primary_preempt_buf_size, secondary_preempt_buf_size;
+
+	if (fw_hdr->preemption_buffer_1_max_size)
+		primary_preempt_buf_size = fw_hdr->preemption_buffer_1_max_size;
+	else
+		primary_preempt_buf_size = fw_hdr->preemption_buffer_1_size;
+
+	if (fw_hdr->preemption_buffer_2_max_size)
+		secondary_preempt_buf_size = fw_hdr->preemption_buffer_2_max_size;
+	else
+		secondary_preempt_buf_size = fw_hdr->preemption_buffer_2_size;
+
+	ivpu_dbg(vdev, FW_BOOT, "Preemption buffer size, primary: %u, secondary: %u\n",
+		 primary_preempt_buf_size, secondary_preempt_buf_size);
+
+	if (primary_preempt_buf_size < FW_PREEMPT_BUF_MIN_SIZE ||
+	    secondary_preempt_buf_size < FW_PREEMPT_BUF_MIN_SIZE) {
+		ivpu_warn(vdev, "Preemption buffers size too small\n");
+		return;
+	}
+
+	if (primary_preempt_buf_size > FW_PREEMPT_BUF_MAX_SIZE ||
+	    secondary_preempt_buf_size > FW_PREEMPT_BUF_MAX_SIZE) {
+		ivpu_warn(vdev, "Preemption buffers size too big\n");
+		return;
+	}
+
+	if (fw->sched_mode != VPU_SCHEDULING_MODE_HW)
+		return;
+
+	if (ivpu_test_mode & IVPU_TEST_MODE_MIP_DISABLE)
+		return;
+
+	vdev->fw->primary_preempt_buf_size = ALIGN(primary_preempt_buf_size, PAGE_SIZE);
+	vdev->fw->secondary_preempt_buf_size = ALIGN(secondary_preempt_buf_size, PAGE_SIZE);
+}
+
 static int ivpu_fw_parse(struct ivpu_device *vdev)
 {
 	struct ivpu_fw_info *fw = vdev->fw;
 	const struct vpu_firmware_header *fw_hdr = (const void *)fw->file->data;
-	u64 runtime_addr, image_load_addr, runtime_size, image_size;
+	struct ivpu_addr_range fw_image_range;
+	u64 boot_params_addr, boot_params_size;
+	u64 fw_version_addr, fw_version_size;
+	u64 runtime_addr, runtime_size;
+	u64 image_load_addr, image_size;
 
 	if (fw->file->size <= FW_FILE_IMAGE_OFFSET) {
 		ivpu_err(vdev, "Firmware file is too small: %zu\n", fw->file->size);
@@ -167,18 +212,37 @@ static int ivpu_fw_parse(struct ivpu_device *vdev)
 		return -EINVAL;
 	}
 
-	runtime_addr = fw_hdr->boot_params_load_address;
-	runtime_size = fw_hdr->runtime_size;
-	image_load_addr = fw_hdr->image_load_address;
-	image_size = fw_hdr->image_size;
+	boot_params_addr = fw_hdr->boot_params_load_address;
+	boot_params_size = SZ_4K;
+
+	if (!ivpu_is_within_range(boot_params_addr, boot_params_size, &vdev->hw->ranges.runtime)) {
+		ivpu_err(vdev, "Invalid boot params address: 0x%llx\n", boot_params_addr);
+		return -EINVAL;
+	}
+
+	fw_version_addr = fw_hdr->firmware_version_load_address;
+	fw_version_size = ALIGN(fw_hdr->firmware_version_size, SZ_4K);
 
-	if (runtime_addr < FW_RUNTIME_MIN_ADDR || runtime_addr > FW_RUNTIME_MAX_ADDR) {
-		ivpu_err(vdev, "Invalid firmware runtime address: 0x%llx\n", runtime_addr);
+	if (fw_version_size != SZ_4K) {
+		ivpu_err(vdev, "Invalid firmware version size: %u\n",
+			 fw_hdr->firmware_version_size);
 		return -EINVAL;
 	}
 
-	if (runtime_size < fw->file->size || runtime_size > FW_RUNTIME_MAX_SIZE) {
-		ivpu_err(vdev, "Invalid firmware runtime size: %llu\n", runtime_size);
+	if (!ivpu_is_within_range(fw_version_addr, fw_version_size, &vdev->hw->ranges.runtime)) {
+		ivpu_err(vdev, "Invalid firmware version address: 0x%llx\n", fw_version_addr);
+		return -EINVAL;
+	}
+
+	runtime_addr = fw_hdr->image_load_address;
+	runtime_size = fw_hdr->runtime_size - boot_params_size - fw_version_size;
+
+	image_load_addr = fw_hdr->image_load_address;
+	image_size = fw_hdr->image_size;
+
+	if (!ivpu_is_within_range(runtime_addr, runtime_size, &vdev->hw->ranges.runtime)) {
+		ivpu_err(vdev, "Invalid firmware runtime address: 0x%llx and size %llu\n",
+			 runtime_addr, runtime_size);
 		return -EINVAL;
 	}
 
@@ -187,23 +251,25 @@ static int ivpu_fw_parse(struct ivpu_device *vdev)
 		return -EINVAL;
 	}
 
-	if (image_load_addr < runtime_addr ||
-	    image_load_addr + image_size > runtime_addr + runtime_size) {
-		ivpu_err(vdev, "Invalid firmware load address size: 0x%llx and size %llu\n",
+	if (!ivpu_is_within_range(image_load_addr, image_size, &vdev->hw->ranges.runtime)) {
+		ivpu_err(vdev, "Invalid firmware load address: 0x%llx and size %llu\n",
 			 image_load_addr, image_size);
 		return -EINVAL;
 	}
 
-	if (fw_hdr->shave_nn_fw_size > FW_SHAVE_NN_MAX_SIZE) {
-		ivpu_err(vdev, "SHAVE NN firmware is too big: %u\n", fw_hdr->shave_nn_fw_size);
+	if (ivpu_hw_range_init(vdev, &fw_image_range, image_load_addr, image_size))
 		return -EINVAL;
-	}
 
-	if (fw_hdr->entry_point < image_load_addr ||
-	    fw_hdr->entry_point >= image_load_addr + image_size) {
+	if (!ivpu_is_within_range(fw_hdr->entry_point, SZ_4K, &fw_image_range)) {
 		ivpu_err(vdev, "Invalid entry point: 0x%llx\n", fw_hdr->entry_point);
 		return -EINVAL;
 	}
+
+	if (fw_hdr->shave_nn_fw_size > FW_SHAVE_NN_MAX_SIZE) {
+		ivpu_err(vdev, "SHAVE NN firmware is too big: %u\n", fw_hdr->shave_nn_fw_size);
+		return -EINVAL;
+	}
+
 	ivpu_dbg(vdev, FW_BOOT, "Header version: 0x%x, format 0x%x\n",
 		 fw_hdr->header_version, fw_hdr->image_format);
 
@@ -217,6 +283,10 @@ static int ivpu_fw_parse(struct ivpu_device *vdev)
 	if (IVPU_FW_CHECK_API_COMPAT(vdev, fw_hdr, JSM, 3))
 		return -EINVAL;
 
+	fw->boot_params_addr = boot_params_addr;
+	fw->boot_params_size = boot_params_size;
+	fw->fw_version_addr = fw_version_addr;
+	fw->fw_version_size = fw_version_size;
 	fw->runtime_addr = runtime_addr;
 	fw->runtime_size = runtime_size;
 	fw->image_load_offset = image_load_addr - runtime_addr;
@@ -235,22 +305,13 @@ static int ivpu_fw_parse(struct ivpu_device *vdev)
 	fw->sched_mode = ivpu_fw_sched_mode_select(vdev, fw_hdr);
 	ivpu_info(vdev, "Scheduler mode: %s\n", fw->sched_mode ? "HW" : "OS");
 
-	if (fw_hdr->preemption_buffer_1_max_size)
-		fw->primary_preempt_buf_size = fw_hdr->preemption_buffer_1_max_size;
-	else
-		fw->primary_preempt_buf_size = fw_hdr->preemption_buffer_1_size;
+	ivpu_preemption_config_parse(vdev, fw_hdr);
+	ivpu_dbg(vdev, FW_BOOT, "Mid-inference preemption %s supported\n",
+		 ivpu_fw_preempt_buf_size(vdev) ? "is" : "is not");
 
-	if (fw_hdr->preemption_buffer_2_max_size)
-		fw->secondary_preempt_buf_size = fw_hdr->preemption_buffer_2_max_size;
-	else
-		fw->secondary_preempt_buf_size = fw_hdr->preemption_buffer_2_size;
-	ivpu_dbg(vdev, FW_BOOT, "Preemption buffer sizes: primary %u, secondary %u\n",
-		 fw->primary_preempt_buf_size, fw->secondary_preempt_buf_size);
-
-	if (fw_hdr->ro_section_start_address && !is_within_range(fw_hdr->ro_section_start_address,
-								 fw_hdr->ro_section_size,
-								 fw_hdr->image_load_address,
-								 fw_hdr->image_size)) {
+	if (fw_hdr->ro_section_start_address &&
+	    !ivpu_is_within_range(fw_hdr->ro_section_start_address, fw_hdr->ro_section_size,
+				  &fw_image_range)) {
 		ivpu_err(vdev, "Invalid read-only section: start address 0x%llx, size %u\n",
 			 fw_hdr->ro_section_start_address, fw_hdr->ro_section_size);
 		return -EINVAL;
@@ -259,12 +320,18 @@ static int ivpu_fw_parse(struct ivpu_device *vdev)
 	fw->read_only_addr = fw_hdr->ro_section_start_address;
 	fw->read_only_size = fw_hdr->ro_section_size;
 
-	ivpu_dbg(vdev, FW_BOOT, "Size: file %lu image %u runtime %u shavenn %u\n",
-		 fw->file->size, fw->image_size, fw->runtime_size, fw->shave_nn_size);
-	ivpu_dbg(vdev, FW_BOOT, "Address: runtime 0x%llx, load 0x%llx, entry point 0x%llx\n",
-		 fw->runtime_addr, image_load_addr, fw->entry_point);
+	ivpu_dbg(vdev, FW_BOOT, "Boot params: address 0x%llx, size %llu\n",
+		 fw->boot_params_addr, fw->boot_params_size);
+	ivpu_dbg(vdev, FW_BOOT, "FW version:  address 0x%llx, size %llu\n",
+		 fw->fw_version_addr, fw->fw_version_size);
+	ivpu_dbg(vdev, FW_BOOT, "Runtime:     address 0x%llx, size %u\n",
+		 fw->runtime_addr, fw->runtime_size);
+	ivpu_dbg(vdev, FW_BOOT, "Image load offset: 0x%llx, size %u\n",
+		 fw->image_load_offset, fw->image_size);
 	ivpu_dbg(vdev, FW_BOOT, "Read-only section: address 0x%llx, size %u\n",
 		 fw->read_only_addr, fw->read_only_size);
+	ivpu_dbg(vdev, FW_BOOT, "FW entry point: 0x%llx\n", fw->entry_point);
+	ivpu_dbg(vdev, FW_BOOT, "SHAVE NN size: %u\n", fw->shave_nn_size);
 
 	return 0;
 }
@@ -291,39 +358,33 @@ ivpu_fw_init_wa(struct ivpu_device *vdev)
 	IVPU_PRINT_WA(disable_d0i3_msg);
 }
 
-static int ivpu_fw_update_global_range(struct ivpu_device *vdev)
-{
-	struct ivpu_fw_info *fw = vdev->fw;
-	u64 start = ALIGN(fw->runtime_addr + fw->runtime_size, FW_SHARED_MEM_ALIGNMENT);
-	u64 size = FW_SHARED_MEM_SIZE;
-
-	if (start + size > FW_GLOBAL_MEM_END) {
-		ivpu_err(vdev, "No space for shared region, start %lld, size %lld\n", start, size);
-		return -EINVAL;
-	}
-
-	ivpu_hw_range_init(&vdev->hw->ranges.global, start, size);
-	return 0;
-}
-
 static int ivpu_fw_mem_init(struct ivpu_device *vdev)
 {
 	struct ivpu_fw_info *fw = vdev->fw;
-	struct ivpu_addr_range fw_range;
 	int log_verb_size;
 	int ret;
 
-	ret = ivpu_fw_update_global_range(vdev);
-	if (ret)
-		return ret;
+	fw->mem_bp = ivpu_bo_create_runtime(vdev, fw->boot_params_addr, fw->boot_params_size,
+					    DRM_IVPU_BO_WC | DRM_IVPU_BO_MAPPABLE);
+	if (!fw->mem_bp) {
+		ivpu_err(vdev, "Failed to create firmware boot params memory buffer\n");
+		return -ENOMEM;
+	}
+
+	fw->mem_fw_ver = ivpu_bo_create_runtime(vdev, fw->fw_version_addr, fw->fw_version_size,
+						DRM_IVPU_BO_WC | DRM_IVPU_BO_MAPPABLE);
+	if (!fw->mem_fw_ver) {
+		ivpu_err(vdev, "Failed to create firmware version memory buffer\n");
+		ret = -ENOMEM;
+		goto err_free_bp;
+	}
 
-	fw_range.start = fw->runtime_addr;
-	fw_range.end = fw->runtime_addr + fw->runtime_size;
-	fw->mem = ivpu_bo_create(vdev, &vdev->gctx, &fw_range, fw->runtime_size,
-				 DRM_IVPU_BO_WC | DRM_IVPU_BO_MAPPABLE);
+	fw->mem = ivpu_bo_create_runtime(vdev, fw->runtime_addr, fw->runtime_size,
+					 DRM_IVPU_BO_WC | DRM_IVPU_BO_MAPPABLE);
 	if (!fw->mem) {
 		ivpu_err(vdev, "Failed to create firmware runtime memory buffer\n");
-		return -ENOMEM;
+		ret = -ENOMEM;
+		goto err_free_fw_ver;
 	}
 
 	ret = ivpu_mmu_context_set_pages_ro(vdev, &vdev->gctx, fw->read_only_addr,
@@ -372,6 +433,10 @@ err_free_log_crit:
 	ivpu_bo_free(fw->mem_log_crit);
 err_free_fw_mem:
 	ivpu_bo_free(fw->mem);
+err_free_fw_ver:
+	ivpu_bo_free(fw->mem_fw_ver);
+err_free_bp:
+	ivpu_bo_free(fw->mem_bp);
 	return ret;
 }
 
@@ -387,10 +452,14 @@ static void ivpu_fw_mem_fini(struct ivpu_device *vdev)
 	ivpu_bo_free(fw->mem_log_verb);
 	ivpu_bo_free(fw->mem_log_crit);
 	ivpu_bo_free(fw->mem);
+	ivpu_bo_free(fw->mem_fw_ver);
+	ivpu_bo_free(fw->mem_bp);
 
 	fw->mem_log_verb = NULL;
 	fw->mem_log_crit = NULL;
 	fw->mem = NULL;
+	fw->mem_fw_ver = NULL;
+	fw->mem_bp = NULL;
 }
 
 int ivpu_fw_init(struct ivpu_device *vdev)
@@ -483,11 +552,6 @@ static void ivpu_fw_boot_params_print(struct ivpu_device *vdev, struct vpu_boot_
 	ivpu_dbg(vdev, FW_BOOT, "boot_params.cache_defaults[VPU_BOOT_L2_CACHE_CFG_NN].cfg = 0x%x\n",
 		 boot_params->cache_defaults[VPU_BOOT_L2_CACHE_CFG_NN].cfg);
 
-	ivpu_dbg(vdev, FW_BOOT, "boot_params.global_memory_allocator_base = 0x%llx\n",
-		 boot_params->global_memory_allocator_base);
-	ivpu_dbg(vdev, FW_BOOT, "boot_params.global_memory_allocator_size = 0x%x\n",
-		 boot_params->global_memory_allocator_size);
-
 	ivpu_dbg(vdev, FW_BOOT, "boot_params.shave_nn_fw_base = 0x%llx\n",
 		 boot_params->shave_nn_fw_base);
 
@@ -495,10 +559,6 @@ static void ivpu_fw_boot_params_print(struct ivpu_device *vdev, struct vpu_boot_
 		 boot_params->watchdog_irq_mss);
 	ivpu_dbg(vdev, FW_BOOT, "boot_params.watchdog_irq_nce = 0x%x\n",
 		 boot_params->watchdog_irq_nce);
-	ivpu_dbg(vdev, FW_BOOT, "boot_params.host_to_vpu_irq = 0x%x\n",
-		 boot_params->host_to_vpu_irq);
-	ivpu_dbg(vdev, FW_BOOT, "boot_params.job_done_irq = 0x%x\n",
-		 boot_params->job_done_irq);
 
 	ivpu_dbg(vdev, FW_BOOT, "boot_params.host_version_id = 0x%x\n",
 		 boot_params->host_version_id);
@@ -546,6 +606,8 @@ static void ivpu_fw_boot_params_print(struct ivpu_device *vdev, struct vpu_boot_
 		 boot_params->system_time_us);
 	ivpu_dbg(vdev, FW_BOOT, "boot_params.power_profile = 0x%x\n",
 		 boot_params->power_profile);
+	ivpu_dbg(vdev, FW_BOOT, "boot_params.vpu_uses_ecc_mca_signal = 0x%x\n",
+		 boot_params->vpu_uses_ecc_mca_signal);
 }
 
 void ivpu_fw_boot_params_setup(struct ivpu_device *vdev, struct vpu_boot_params *boot_params)
@@ -572,6 +634,7 @@ void ivpu_fw_boot_params_setup(struct ivpu_device *vdev, struct vpu_boot_params
 		return;
 	}
 
+	memset(boot_params, 0, sizeof(*boot_params));
 	vdev->pm->is_warmboot = false;
 
 	boot_params->magic = VPU_BOOT_PARAMS_MAGIC;
@@ -647,6 +710,8 @@ void ivpu_fw_boot_params_setup(struct ivpu_device *vdev, struct vpu_boot_params
 	boot_params->d0i3_entry_vpu_ts = 0;
 	if (IVPU_WA(disable_d0i2))
 		boot_params->power_profile |= BIT(1);
+	boot_params->vpu_uses_ecc_mca_signal =
+		ivpu_hw_uses_ecc_mca_signal(vdev) ? VPU_BOOT_MCA_ECC_BOTH : 0;
 
 	boot_params->system_time_us = ktime_to_us(ktime_get_real());
 	wmb(); /* Flush WC buffers after writing bootparams */
diff --git a/drivers/accel/ivpu/ivpu_fw.h b/drivers/accel/ivpu/ivpu_fw.h
index 7081913fb0dd..00945892b55e 100644
--- a/drivers/accel/ivpu/ivpu_fw.h
+++ b/drivers/accel/ivpu/ivpu_fw.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0-only */
 /*
- * Copyright (C) 2020-2024 Intel Corporation
+ * Copyright (C) 2020-2025 Intel Corporation
  */
 
 #ifndef __IVPU_FW_H__
@@ -19,10 +19,16 @@ struct ivpu_fw_info {
 	const struct firmware *file;
 	const char *name;
 	char version[FW_VERSION_STR_SIZE];
+	struct ivpu_bo *mem_bp;
+	struct ivpu_bo *mem_fw_ver;
 	struct ivpu_bo *mem;
 	struct ivpu_bo *mem_shave_nn;
 	struct ivpu_bo *mem_log_crit;
 	struct ivpu_bo *mem_log_verb;
+	u64 boot_params_addr;
+	u64 boot_params_size;
+	u64 fw_version_addr;
+	u64 fw_version_size;
 	u64 runtime_addr;
 	u32 runtime_size;
 	u64 image_load_offset;
@@ -42,6 +48,7 @@ struct ivpu_fw_info {
 	u64 last_heartbeat;
 };
 
+bool ivpu_is_within_range(u64 addr, size_t size, struct ivpu_addr_range *range);
 int ivpu_fw_init(struct ivpu_device *vdev);
 void ivpu_fw_fini(struct ivpu_device *vdev);
 void ivpu_fw_load(struct ivpu_device *vdev);
@@ -52,4 +59,9 @@ static inline bool ivpu_fw_is_cold_boot(struct ivpu_device *vdev)
 	return vdev->fw->entry_point == vdev->fw->cold_boot_entry_point;
 }
 
+static inline u32 ivpu_fw_preempt_buf_size(struct ivpu_device *vdev)
+{
+	return vdev->fw->primary_preempt_buf_size + vdev->fw->secondary_preempt_buf_size;
+}
+
 #endif /* __IVPU_FW_H__ */
diff --git a/drivers/accel/ivpu/ivpu_gem.c b/drivers/accel/ivpu/ivpu_gem.c
index 59cfcf3eaded..e9830ad48d4b 100644
--- a/drivers/accel/ivpu/ivpu_gem.c
+++ b/drivers/accel/ivpu/ivpu_gem.c
@@ -15,6 +15,7 @@
 #include <drm/drm_utils.h>
 
 #include "ivpu_drv.h"
+#include "ivpu_fw.h"
 #include "ivpu_gem.h"
 #include "ivpu_hw.h"
 #include "ivpu_mmu.h"
@@ -27,8 +28,8 @@ static const struct drm_gem_object_funcs ivpu_gem_funcs;
 static inline void ivpu_dbg_bo(struct ivpu_device *vdev, struct ivpu_bo *bo, const char *action)
 {
 	ivpu_dbg(vdev, BO,
-		 "%6s: bo %8p vpu_addr %9llx size %8zu ctx %d has_pages %d dma_mapped %d mmu_mapped %d wc %d imported %d\n",
-		 action, bo, bo->vpu_addr, ivpu_bo_size(bo), bo->ctx_id,
+		 "%6s: bo %8p size %9zu ctx %d vpu_addr %9llx pages %d sgt %d mmu_mapped %d wc %d imported %d\n",
+		 action, bo, ivpu_bo_size(bo), bo->ctx_id, bo->vpu_addr,
 		 (bool)bo->base.pages, (bool)bo->base.sgt, bo->mmu_mapped, bo->base.map_wc,
 		 (bool)drm_gem_is_imported(&bo->base.base));
 }
@@ -43,22 +44,46 @@ static inline void ivpu_bo_unlock(struct ivpu_bo *bo)
 	dma_resv_unlock(bo->base.base.resv);
 }
 
+static struct sg_table *ivpu_bo_map_attachment(struct ivpu_device *vdev, struct ivpu_bo *bo)
+{
+	struct sg_table *sgt = bo->base.sgt;
+
+	drm_WARN_ON(&vdev->drm, !bo->base.base.import_attach);
+
+	ivpu_bo_lock(bo);
+
+	if (!sgt) {
+		sgt = dma_buf_map_attachment(bo->base.base.import_attach, DMA_BIDIRECTIONAL);
+		if (IS_ERR(sgt))
+			ivpu_err(vdev, "Failed to map BO in IOMMU: %ld\n", PTR_ERR(sgt));
+		else
+			bo->base.sgt = sgt;
+	}
+
+	ivpu_bo_unlock(bo);
+
+	return sgt;
+}
+
 /*
- * ivpu_bo_pin() - pin the backing physical pages and map them to VPU.
+ * ivpu_bo_bind() - pin the backing physical pages and map them to VPU.
  *
  * This function pins physical memory pages, then maps the physical pages
  * to IOMMU address space and finally updates the VPU MMU page tables
  * to allow the VPU to translate VPU address to IOMMU address.
  */
-int __must_check ivpu_bo_pin(struct ivpu_bo *bo)
+int __must_check ivpu_bo_bind(struct ivpu_bo *bo)
 {
 	struct ivpu_device *vdev = ivpu_bo_to_vdev(bo);
 	struct sg_table *sgt;
 	int ret = 0;
 
-	ivpu_dbg_bo(vdev, bo, "pin");
+	ivpu_dbg_bo(vdev, bo, "bind");
 
-	sgt = drm_gem_shmem_get_pages_sgt(&bo->base);
+	if (bo->base.base.import_attach)
+		sgt = ivpu_bo_map_attachment(vdev, bo);
+	else
+		sgt = drm_gem_shmem_get_pages_sgt(&bo->base);
 	if (IS_ERR(sgt)) {
 		ret = PTR_ERR(sgt);
 		ivpu_err(vdev, "Failed to map BO in IOMMU: %d\n", ret);
@@ -99,7 +124,9 @@ ivpu_bo_alloc_vpu_addr(struct ivpu_bo *bo, struct ivpu_mmu_context *ctx,
 	ret = ivpu_mmu_context_insert_node(ctx, range, ivpu_bo_size(bo), &bo->mm_node);
 	if (!ret) {
 		bo->ctx = ctx;
+		bo->ctx_id = ctx->id;
 		bo->vpu_addr = bo->mm_node.start;
+		ivpu_dbg_bo(vdev, bo, "vaddr");
 	} else {
 		ivpu_err(vdev, "Failed to add BO to context %u: %d\n", ctx->id, ret);
 	}
@@ -115,7 +142,7 @@ static void ivpu_bo_unbind_locked(struct ivpu_bo *bo)
 {
 	struct ivpu_device *vdev = ivpu_bo_to_vdev(bo);
 
-	lockdep_assert(dma_resv_held(bo->base.base.resv) || !kref_read(&bo->base.base.refcount));
+	dma_resv_assert_held(bo->base.base.resv);
 
 	if (bo->mmu_mapped) {
 		drm_WARN_ON(&vdev->drm, !bo->ctx);
@@ -134,9 +161,14 @@ static void ivpu_bo_unbind_locked(struct ivpu_bo *bo)
 		return;
 
 	if (bo->base.sgt) {
-		dma_unmap_sgtable(vdev->drm.dev, bo->base.sgt, DMA_BIDIRECTIONAL, 0);
-		sg_free_table(bo->base.sgt);
-		kfree(bo->base.sgt);
+		if (bo->base.base.import_attach) {
+			dma_buf_unmap_attachment(bo->base.base.import_attach,
+						 bo->base.sgt, DMA_BIDIRECTIONAL);
+		} else {
+			dma_unmap_sgtable(vdev->drm.dev, bo->base.sgt, DMA_BIDIRECTIONAL, 0);
+			sg_free_table(bo->base.sgt);
+			kfree(bo->base.sgt);
+		}
 		bo->base.sgt = NULL;
 	}
 }
@@ -182,10 +214,11 @@ struct drm_gem_object *ivpu_gem_create_object(struct drm_device *dev, size_t siz
 struct drm_gem_object *ivpu_gem_prime_import(struct drm_device *dev,
 					     struct dma_buf *dma_buf)
 {
+	struct ivpu_device *vdev = to_ivpu_device(dev);
 	struct device *attach_dev = dev->dev;
 	struct dma_buf_attachment *attach;
-	struct sg_table *sgt;
 	struct drm_gem_object *obj;
+	struct ivpu_bo *bo;
 	int ret;
 
 	attach = dma_buf_attach(dma_buf, attach_dev);
@@ -194,25 +227,25 @@ struct drm_gem_object *ivpu_gem_prime_import(struct drm_device *dev,
 
 	get_dma_buf(dma_buf);
 
-	sgt = dma_buf_map_attachment_unlocked(attach, DMA_BIDIRECTIONAL);
-	if (IS_ERR(sgt)) {
-		ret = PTR_ERR(sgt);
-		goto fail_detach;
-	}
-
-	obj = drm_gem_shmem_prime_import_sg_table(dev, attach, sgt);
+	obj = drm_gem_shmem_prime_import_sg_table(dev, attach, NULL);
 	if (IS_ERR(obj)) {
 		ret = PTR_ERR(obj);
-		goto fail_unmap;
+		goto fail_detach;
 	}
 
 	obj->import_attach = attach;
 	obj->resv = dma_buf->resv;
 
+	bo = to_ivpu_bo(obj);
+
+	mutex_lock(&vdev->bo_list_lock);
+	list_add_tail(&bo->bo_list_node, &vdev->bo_list);
+	mutex_unlock(&vdev->bo_list_lock);
+
+	ivpu_dbg(vdev, BO, "import: bo %8p size %9zu\n", bo, ivpu_bo_size(bo));
+
 	return obj;
 
-fail_unmap:
-	dma_buf_unmap_attachment_unlocked(attach, sgt, DMA_BIDIRECTIONAL);
 fail_detach:
 	dma_buf_detach(dma_buf, attach);
 	dma_buf_put(dma_buf);
@@ -220,7 +253,7 @@ fail_detach:
 	return ERR_PTR(ret);
 }
 
-static struct ivpu_bo *ivpu_bo_alloc(struct ivpu_device *vdev, u64 size, u32 flags, u32 ctx_id)
+static struct ivpu_bo *ivpu_bo_alloc(struct ivpu_device *vdev, u64 size, u32 flags)
 {
 	struct drm_gem_shmem_object *shmem;
 	struct ivpu_bo *bo;
@@ -238,7 +271,6 @@ static struct ivpu_bo *ivpu_bo_alloc(struct ivpu_device *vdev, u64 size, u32 fla
 		return ERR_CAST(shmem);
 
 	bo = to_ivpu_bo(&shmem->base);
-	bo->ctx_id = ctx_id;
 	bo->base.map_wc = flags & DRM_IVPU_BO_WC;
 	bo->flags = flags;
 
@@ -246,7 +278,7 @@ static struct ivpu_bo *ivpu_bo_alloc(struct ivpu_device *vdev, u64 size, u32 fla
 	list_add_tail(&bo->bo_list_node, &vdev->bo_list);
 	mutex_unlock(&vdev->bo_list_lock);
 
-	ivpu_dbg_bo(vdev, bo, "alloc");
+	ivpu_dbg(vdev, BO, " alloc: bo %8p size %9llu\n", bo, size);
 
 	return bo;
 }
@@ -281,6 +313,8 @@ static void ivpu_gem_bo_free(struct drm_gem_object *obj)
 
 	ivpu_dbg_bo(vdev, bo, "free");
 
+	drm_WARN_ON(&vdev->drm, list_empty(&bo->bo_list_node));
+
 	mutex_lock(&vdev->bo_list_lock);
 	list_del(&bo->bo_list_node);
 	mutex_unlock(&vdev->bo_list_lock);
@@ -290,11 +324,15 @@ static void ivpu_gem_bo_free(struct drm_gem_object *obj)
 	drm_WARN_ON(&vdev->drm, ivpu_bo_size(bo) == 0);
 	drm_WARN_ON(&vdev->drm, bo->base.vaddr);
 
+	ivpu_bo_lock(bo);
 	ivpu_bo_unbind_locked(bo);
+	ivpu_bo_unlock(bo);
+
 	drm_WARN_ON(&vdev->drm, bo->mmu_mapped);
 	drm_WARN_ON(&vdev->drm, bo->ctx);
 
 	drm_WARN_ON(obj->dev, refcount_read(&bo->base.pages_use_count) > 1);
+	drm_WARN_ON(obj->dev, bo->base.base.vma_node.vm_files.rb_node);
 	drm_gem_shmem_free(&bo->base);
 }
 
@@ -326,19 +364,23 @@ int ivpu_bo_create_ioctl(struct drm_device *dev, void *data, struct drm_file *fi
 	if (size == 0)
 		return -EINVAL;
 
-	bo = ivpu_bo_alloc(vdev, size, args->flags, file_priv->ctx.id);
+	bo = ivpu_bo_alloc(vdev, size, args->flags);
 	if (IS_ERR(bo)) {
 		ivpu_err(vdev, "Failed to allocate BO: %pe (ctx %u size %llu flags 0x%x)",
 			 bo, file_priv->ctx.id, args->size, args->flags);
 		return PTR_ERR(bo);
 	}
 
+	drm_WARN_ON(&vdev->drm, bo->base.base.handle_count != 0);
+
 	ret = drm_gem_handle_create(file, &bo->base.base, &args->handle);
-	if (ret)
+	if (ret) {
 		ivpu_err(vdev, "Failed to create handle for BO: %pe (ctx %u size %llu flags 0x%x)",
 			 bo, file_priv->ctx.id, args->size, args->flags);
-	else
+	} else {
 		args->vpu_addr = bo->vpu_addr;
+		drm_WARN_ON(&vdev->drm, bo->base.base.handle_count != 1);
+	}
 
 	drm_gem_object_put(&bo->base.base);
 
@@ -360,7 +402,7 @@ ivpu_bo_create(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx,
 	drm_WARN_ON(&vdev->drm, !PAGE_ALIGNED(range->end));
 	drm_WARN_ON(&vdev->drm, !PAGE_ALIGNED(size));
 
-	bo = ivpu_bo_alloc(vdev, size, flags, IVPU_GLOBAL_CONTEXT_MMU_SSID);
+	bo = ivpu_bo_alloc(vdev, size, flags);
 	if (IS_ERR(bo)) {
 		ivpu_err(vdev, "Failed to allocate BO: %pe (vpu_addr 0x%llx size %llu flags 0x%x)",
 			 bo, range->start, size, flags);
@@ -371,7 +413,7 @@ ivpu_bo_create(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx,
 	if (ret)
 		goto err_put;
 
-	ret = ivpu_bo_pin(bo);
+	ret = ivpu_bo_bind(bo);
 	if (ret)
 		goto err_put;
 
@@ -391,6 +433,21 @@ err_put:
 	return NULL;
 }
 
+struct ivpu_bo *ivpu_bo_create_runtime(struct ivpu_device *vdev, u64 addr, u64 size, u32 flags)
+{
+	struct ivpu_addr_range range;
+
+	if (!ivpu_is_within_range(addr, size, &vdev->hw->ranges.runtime)) {
+		ivpu_err(vdev, "Invalid runtime BO address 0x%llx size %llu\n", addr, size);
+		return NULL;
+	}
+
+	if (ivpu_hw_range_init(vdev, &range, addr, size))
+		return NULL;
+
+	return ivpu_bo_create(vdev, &vdev->gctx, &range, size, flags);
+}
+
 struct ivpu_bo *ivpu_bo_create_global(struct ivpu_device *vdev, u64 size, u32 flags)
 {
 	return ivpu_bo_create(vdev, &vdev->gctx, &vdev->hw->ranges.global, size, flags);
diff --git a/drivers/accel/ivpu/ivpu_gem.h b/drivers/accel/ivpu/ivpu_gem.h
index aa8ff14f7aae..54452eb8a41f 100644
--- a/drivers/accel/ivpu/ivpu_gem.h
+++ b/drivers/accel/ivpu/ivpu_gem.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0-only */
 /*
- * Copyright (C) 2020-2023 Intel Corporation
+ * Copyright (C) 2020-2025 Intel Corporation
  */
 #ifndef __IVPU_GEM_H__
 #define __IVPU_GEM_H__
@@ -24,13 +24,14 @@ struct ivpu_bo {
 	bool mmu_mapped;
 };
 
-int ivpu_bo_pin(struct ivpu_bo *bo);
+int ivpu_bo_bind(struct ivpu_bo *bo);
 void ivpu_bo_unbind_all_bos_from_context(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx);
 
 struct drm_gem_object *ivpu_gem_create_object(struct drm_device *dev, size_t size);
 struct drm_gem_object *ivpu_gem_prime_import(struct drm_device *dev, struct dma_buf *dma_buf);
 struct ivpu_bo *ivpu_bo_create(struct ivpu_device *vdev, struct ivpu_mmu_context *ctx,
 			       struct ivpu_addr_range *range, u64 size, u32 flags);
+struct ivpu_bo *ivpu_bo_create_runtime(struct ivpu_device *vdev, u64 addr, u64 size, u32 flags);
 struct ivpu_bo *ivpu_bo_create_global(struct ivpu_device *vdev, u64 size, u32 flags);
 void ivpu_bo_free(struct ivpu_bo *bo);
 
@@ -96,4 +97,9 @@ static inline u32 cpu_to_vpu_addr(struct ivpu_bo *bo, void *cpu_addr)
 	return bo->vpu_addr + (cpu_addr - ivpu_bo_vaddr(bo));
 }
 
+static inline bool ivpu_bo_is_mappable(struct ivpu_bo *bo)
+{
+	return bo->flags & DRM_IVPU_BO_MAPPABLE;
+}
+
 #endif /* __IVPU_GEM_H__ */
diff --git a/drivers/accel/ivpu/ivpu_hw.c b/drivers/accel/ivpu/ivpu_hw.c
index 08dcc31b56f4..d69cd0d93569 100644
--- a/drivers/accel/ivpu/ivpu_hw.c
+++ b/drivers/accel/ivpu/ivpu_hw.c
@@ -8,6 +8,8 @@
 #include "ivpu_hw_btrs.h"
 #include "ivpu_hw_ip.h"
 
+#include <asm/msr-index.h>
+#include <asm/msr.h>
 #include <linux/dmi.h>
 #include <linux/fault-inject.h>
 #include <linux/pm_runtime.h>
@@ -20,6 +22,10 @@ module_param_named_unsafe(fail_hw, ivpu_fail_hw, charp, 0444);
 MODULE_PARM_DESC(fail_hw, "<interval>,<probability>,<space>,<times>");
 #endif
 
+#define FW_SHARED_MEM_ALIGNMENT	SZ_512K /* VPU MTRR limitation */
+
+#define ECC_MCA_SIGNAL_ENABLE_MASK	0xff
+
 static char *platform_to_str(u32 platform)
 {
 	switch (platform) {
@@ -147,19 +153,39 @@ static void priority_bands_init(struct ivpu_device *vdev)
 	vdev->hw->hws.process_quantum[VPU_JOB_SCHEDULING_PRIORITY_BAND_REALTIME] = 200000;
 }
 
+int ivpu_hw_range_init(struct ivpu_device *vdev, struct ivpu_addr_range *range, u64 start, u64 size)
+{
+	u64 end;
+
+	if (!range || check_add_overflow(start, size, &end)) {
+		ivpu_err(vdev, "Invalid range: start 0x%llx size %llu\n", start, size);
+		return -EINVAL;
+	}
+
+	range->start = start;
+	range->end = end;
+
+	return 0;
+}
+
 static void memory_ranges_init(struct ivpu_device *vdev)
 {
 	if (ivpu_hw_ip_gen(vdev) == IVPU_HW_IP_37XX) {
-		ivpu_hw_range_init(&vdev->hw->ranges.global, 0x80000000, SZ_512M);
-		ivpu_hw_range_init(&vdev->hw->ranges.user,   0x88000000, 511 * SZ_1M);
-		ivpu_hw_range_init(&vdev->hw->ranges.shave, 0x180000000, SZ_2G);
-		ivpu_hw_range_init(&vdev->hw->ranges.dma,   0x200000000, SZ_128G);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.runtime, 0x84800000, SZ_64M);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.global,  0x90000000, SZ_256M);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.user,    0xa0000000, 511 * SZ_1M);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.shave,  0x180000000, SZ_2G);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.dma,    0x200000000, SZ_128G);
 	} else {
-		ivpu_hw_range_init(&vdev->hw->ranges.global, 0x80000000, SZ_512M);
-		ivpu_hw_range_init(&vdev->hw->ranges.shave,  0x80000000, SZ_2G);
-		ivpu_hw_range_init(&vdev->hw->ranges.user,  0x100000000, SZ_256G);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.runtime, 0x80000000, SZ_64M);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.global,  0x90000000, SZ_256M);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.shave,   0x80000000, SZ_2G);
+		ivpu_hw_range_init(vdev, &vdev->hw->ranges.user,   0x100000000, SZ_256G);
 		vdev->hw->ranges.dma = vdev->hw->ranges.user;
 	}
+
+	drm_WARN_ON(&vdev->drm, !IS_ALIGNED(vdev->hw->ranges.global.start,
+					    FW_SHARED_MEM_ALIGNMENT));
 }
 
 static int wp_enable(struct ivpu_device *vdev)
@@ -373,3 +399,22 @@ irqreturn_t ivpu_hw_irq_handler(int irq, void *ptr)
 	pm_runtime_mark_last_busy(vdev->drm.dev);
 	return IRQ_HANDLED;
 }
+
+bool ivpu_hw_uses_ecc_mca_signal(struct ivpu_device *vdev)
+{
+	unsigned long long msr_integrity_caps;
+	int ret;
+
+	if (ivpu_hw_ip_gen(vdev) < IVPU_HW_IP_50XX)
+		return false;
+
+	ret = rdmsrq_safe(MSR_INTEGRITY_CAPS, &msr_integrity_caps);
+	if (ret) {
+		ivpu_warn(vdev, "Error reading MSR_INTEGRITY_CAPS: %d", ret);
+		return false;
+	}
+
+	ivpu_dbg(vdev, MISC, "MSR_INTEGRITY_CAPS: 0x%llx\n", msr_integrity_caps);
+
+	return msr_integrity_caps & ECC_MCA_SIGNAL_ENABLE_MASK;
+}
diff --git a/drivers/accel/ivpu/ivpu_hw.h b/drivers/accel/ivpu/ivpu_hw.h
index d79668fe1609..b6d0f0d0dccc 100644
--- a/drivers/accel/ivpu/ivpu_hw.h
+++ b/drivers/accel/ivpu/ivpu_hw.h
@@ -21,6 +21,7 @@ struct ivpu_hw_info {
 		bool (*ip_irq_handler)(struct ivpu_device *vdev, int irq);
 	} irq;
 	struct {
+		struct ivpu_addr_range runtime;
 		struct ivpu_addr_range global;
 		struct ivpu_addr_range user;
 		struct ivpu_addr_range shave;
@@ -51,6 +52,8 @@ struct ivpu_hw_info {
 };
 
 int ivpu_hw_init(struct ivpu_device *vdev);
+int ivpu_hw_range_init(struct ivpu_device *vdev, struct ivpu_addr_range *range, u64 start,
+		       u64 size);
 int ivpu_hw_power_up(struct ivpu_device *vdev);
 int ivpu_hw_power_down(struct ivpu_device *vdev);
 int ivpu_hw_reset(struct ivpu_device *vdev);
@@ -60,6 +63,7 @@ void ivpu_irq_handlers_init(struct ivpu_device *vdev);
 void ivpu_hw_irq_enable(struct ivpu_device *vdev);
 void ivpu_hw_irq_disable(struct ivpu_device *vdev);
 irqreturn_t ivpu_hw_irq_handler(int irq, void *ptr);
+bool ivpu_hw_uses_ecc_mca_signal(struct ivpu_device *vdev);
 
 static inline u32 ivpu_hw_btrs_irq_handler(struct ivpu_device *vdev, int irq)
 {
@@ -71,12 +75,6 @@ static inline u32 ivpu_hw_ip_irq_handler(struct ivpu_device *vdev, int irq)
 	return vdev->hw->irq.ip_irq_handler(vdev, irq);
 }
 
-static inline void ivpu_hw_range_init(struct ivpu_addr_range *range, u64 start, u64 size)
-{
-	range->start = start;
-	range->end = start + size;
-}
-
 static inline u64 ivpu_hw_range_size(const struct ivpu_addr_range *range)
 {
 	return range->end - range->start;
diff --git a/drivers/accel/ivpu/ivpu_hw_btrs.c b/drivers/accel/ivpu/ivpu_hw_btrs.c
index afdb3b2aa72a..aa33f562d29c 100644
--- a/drivers/accel/ivpu/ivpu_hw_btrs.c
+++ b/drivers/accel/ivpu/ivpu_hw_btrs.c
@@ -752,7 +752,7 @@ int ivpu_hw_btrs_dct_get_request(struct ivpu_device *vdev, bool *enable)
 	}
 }
 
-void ivpu_hw_btrs_dct_set_status(struct ivpu_device *vdev, bool enable, u32 active_percent)
+void ivpu_hw_btrs_dct_set_status(struct ivpu_device *vdev, bool enable, u8 active_percent)
 {
 	u32 val = 0;
 	u32 cmd = enable ? DCT_ENABLE : DCT_DISABLE;
diff --git a/drivers/accel/ivpu/ivpu_hw_btrs.h b/drivers/accel/ivpu/ivpu_hw_btrs.h
index 032c384ac3d4..c4c10e22f30f 100644
--- a/drivers/accel/ivpu/ivpu_hw_btrs.h
+++ b/drivers/accel/ivpu/ivpu_hw_btrs.h
@@ -36,7 +36,7 @@ u32 ivpu_hw_btrs_dpu_freq_get(struct ivpu_device *vdev);
 bool ivpu_hw_btrs_irq_handler_mtl(struct ivpu_device *vdev, int irq);
 bool ivpu_hw_btrs_irq_handler_lnl(struct ivpu_device *vdev, int irq);
 int ivpu_hw_btrs_dct_get_request(struct ivpu_device *vdev, bool *enable);
-void ivpu_hw_btrs_dct_set_status(struct ivpu_device *vdev, bool enable, u32 active_percent);
+void ivpu_hw_btrs_dct_set_status(struct ivpu_device *vdev, bool enable, u8 active_percent);
 u32 ivpu_hw_btrs_telemetry_offset_get(struct ivpu_device *vdev);
 u32 ivpu_hw_btrs_telemetry_size_get(struct ivpu_device *vdev);
 u32 ivpu_hw_btrs_telemetry_enable_get(struct ivpu_device *vdev);
diff --git a/drivers/accel/ivpu/ivpu_job.c b/drivers/accel/ivpu/ivpu_job.c
index 060f1fc031d3..17273c68f84c 100644
--- a/drivers/accel/ivpu/ivpu_job.c
+++ b/drivers/accel/ivpu/ivpu_job.c
@@ -34,22 +34,20 @@ static void ivpu_cmdq_ring_db(struct ivpu_device *vdev, struct ivpu_cmdq *cmdq)
 static int ivpu_preemption_buffers_create(struct ivpu_device *vdev,
 					  struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cmdq)
 {
-	u64 primary_size = ALIGN(vdev->fw->primary_preempt_buf_size, PAGE_SIZE);
-	u64 secondary_size = ALIGN(vdev->fw->secondary_preempt_buf_size, PAGE_SIZE);
-
-	if (vdev->fw->sched_mode != VPU_SCHEDULING_MODE_HW ||
-	    ivpu_test_mode & IVPU_TEST_MODE_MIP_DISABLE)
+	if (ivpu_fw_preempt_buf_size(vdev) == 0)
 		return 0;
 
 	cmdq->primary_preempt_buf = ivpu_bo_create(vdev, &file_priv->ctx, &vdev->hw->ranges.user,
-						   primary_size, DRM_IVPU_BO_WC);
+						   vdev->fw->primary_preempt_buf_size,
+						   DRM_IVPU_BO_WC);
 	if (!cmdq->primary_preempt_buf) {
 		ivpu_err(vdev, "Failed to create primary preemption buffer\n");
 		return -ENOMEM;
 	}
 
 	cmdq->secondary_preempt_buf = ivpu_bo_create(vdev, &file_priv->ctx, &vdev->hw->ranges.dma,
-						     secondary_size, DRM_IVPU_BO_WC);
+						     vdev->fw->secondary_preempt_buf_size,
+						     DRM_IVPU_BO_WC);
 	if (!cmdq->secondary_preempt_buf) {
 		ivpu_err(vdev, "Failed to create secondary preemption buffer\n");
 		goto err_free_primary;
@@ -66,20 +64,39 @@ err_free_primary:
 static void ivpu_preemption_buffers_free(struct ivpu_device *vdev,
 					 struct ivpu_file_priv *file_priv, struct ivpu_cmdq *cmdq)
 {
-	if (vdev->fw->sched_mode != VPU_SCHEDULING_MODE_HW)
-		return;
-
 	if (cmdq->primary_preempt_buf)
 		ivpu_bo_free(cmdq->primary_preempt_buf);
 	if (cmdq->secondary_preempt_buf)
 		ivpu_bo_free(cmdq->secondary_preempt_buf);
 }
 
+static int ivpu_preemption_job_init(struct ivpu_device *vdev, struct ivpu_file_priv *file_priv,
+				    struct ivpu_cmdq *cmdq, struct ivpu_job *job)
+{
+	int ret;
+
+	/* Use preemption buffer provided by the user space */
+	if (job->primary_preempt_buf)
+		return 0;
+
+	if (!cmdq->primary_preempt_buf) {
+		/* Allocate per command queue preemption buffers */
+		ret = ivpu_preemption_buffers_create(vdev, file_priv, cmdq);
+		if (ret)
+			return ret;
+	}
+
+	/* Use preemption buffers allocated by the kernel */
+	job->primary_preempt_buf = cmdq->primary_preempt_buf;
+	job->secondary_preempt_buf = cmdq->secondary_preempt_buf;
+
+	return 0;
+}
+
 static struct ivpu_cmdq *ivpu_cmdq_alloc(struct ivpu_file_priv *file_priv)
 {
 	struct ivpu_device *vdev = file_priv->vdev;
 	struct ivpu_cmdq *cmdq;
-	int ret;
 
 	cmdq = kzalloc(sizeof(*cmdq), GFP_KERNEL);
 	if (!cmdq)
@@ -89,10 +106,6 @@ static struct ivpu_cmdq *ivpu_cmdq_alloc(struct ivpu_file_priv *file_priv)
 	if (!cmdq->mem)
 		goto err_free_cmdq;
 
-	ret = ivpu_preemption_buffers_create(vdev, file_priv, cmdq);
-	if (ret)
-		ivpu_warn(vdev, "Failed to allocate preemption buffers, preemption limited\n");
-
 	return cmdq;
 
 err_free_cmdq:
@@ -219,11 +232,13 @@ static int ivpu_register_db(struct ivpu_file_priv *file_priv, struct ivpu_cmdq *
 		ret = ivpu_jsm_register_db(vdev, file_priv->ctx.id, cmdq->db_id,
 					   cmdq->mem->vpu_addr, ivpu_bo_size(cmdq->mem));
 
-	if (!ret)
+	if (!ret) {
 		ivpu_dbg(vdev, JOB, "DB %d registered to cmdq %d ctx %d priority %d\n",
 			 cmdq->db_id, cmdq->id, file_priv->ctx.id, cmdq->priority);
-	else
+	} else {
 		xa_erase(&vdev->db_xa, cmdq->db_id);
+		cmdq->db_id = 0;
+	}
 
 	return ret;
 }
@@ -427,17 +442,14 @@ static int ivpu_cmdq_push_job(struct ivpu_cmdq *cmdq, struct ivpu_job *job)
 	if (unlikely(ivpu_test_mode & IVPU_TEST_MODE_NULL_SUBMISSION))
 		entry->flags = VPU_JOB_FLAGS_NULL_SUBMISSION_MASK;
 
-	if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW) {
-		if (cmdq->primary_preempt_buf) {
-			entry->primary_preempt_buf_addr = cmdq->primary_preempt_buf->vpu_addr;
-			entry->primary_preempt_buf_size = ivpu_bo_size(cmdq->primary_preempt_buf);
-		}
+	if (job->primary_preempt_buf) {
+		entry->primary_preempt_buf_addr = job->primary_preempt_buf->vpu_addr;
+		entry->primary_preempt_buf_size = ivpu_bo_size(job->primary_preempt_buf);
+	}
 
-		if (cmdq->secondary_preempt_buf) {
-			entry->secondary_preempt_buf_addr = cmdq->secondary_preempt_buf->vpu_addr;
-			entry->secondary_preempt_buf_size =
-				ivpu_bo_size(cmdq->secondary_preempt_buf);
-		}
+	if (job->secondary_preempt_buf) {
+		entry->secondary_preempt_buf_addr = job->secondary_preempt_buf->vpu_addr;
+		entry->secondary_preempt_buf_size = ivpu_bo_size(job->secondary_preempt_buf);
 	}
 
 	wmb(); /* Ensure that tail is updated after filling entry */
@@ -661,6 +673,13 @@ static int ivpu_job_submit(struct ivpu_job *job, u8 priority, u32 cmdq_id)
 		goto err_unlock;
 	}
 
+	ret = ivpu_preemption_job_init(vdev, file_priv, cmdq, job);
+	if (ret) {
+		ivpu_err(vdev, "Failed to initialize preemption buffers for job %d: %d\n",
+			 job->job_id, ret);
+		goto err_unlock;
+	}
+
 	job->cmdq_id = cmdq->id;
 
 	is_first_job = xa_empty(&vdev->submitted_jobs_xa);
@@ -714,7 +733,7 @@ err_unlock:
 
 static int
 ivpu_job_prepare_bos_for_submit(struct drm_file *file, struct ivpu_job *job, u32 *buf_handles,
-				u32 buf_count, u32 commands_offset)
+				u32 buf_count, u32 commands_offset, u32 preempt_buffer_index)
 {
 	struct ivpu_file_priv *file_priv = job->file_priv;
 	struct ivpu_device *vdev = file_priv->vdev;
@@ -732,7 +751,7 @@ ivpu_job_prepare_bos_for_submit(struct drm_file *file, struct ivpu_job *job, u32
 
 		job->bos[i] = to_ivpu_bo(obj);
 
-		ret = ivpu_bo_pin(job->bos[i]);
+		ret = ivpu_bo_bind(job->bos[i]);
 		if (ret)
 			return ret;
 	}
@@ -750,6 +769,20 @@ ivpu_job_prepare_bos_for_submit(struct drm_file *file, struct ivpu_job *job, u32
 
 	job->cmd_buf_vpu_addr = bo->vpu_addr + commands_offset;
 
+	if (preempt_buffer_index) {
+		struct ivpu_bo *preempt_bo = job->bos[preempt_buffer_index];
+
+		if (ivpu_bo_size(preempt_bo) < ivpu_fw_preempt_buf_size(vdev)) {
+			ivpu_warn(vdev, "Preemption buffer is too small\n");
+			return -EINVAL;
+		}
+		if (ivpu_bo_is_mappable(preempt_bo)) {
+			ivpu_warn(vdev, "Preemption buffer cannot be mappable\n");
+			return -EINVAL;
+		}
+		job->primary_preempt_buf = preempt_bo;
+	}
+
 	ret = drm_gem_lock_reservations((struct drm_gem_object **)job->bos, buf_count,
 					&acquire_ctx);
 	if (ret) {
@@ -780,7 +813,7 @@ unlock_reservations:
 
 static int ivpu_submit(struct drm_file *file, struct ivpu_file_priv *file_priv, u32 cmdq_id,
 		       u32 buffer_count, u32 engine, void __user *buffers_ptr, u32 cmds_offset,
-		       u8 priority)
+		       u32 preempt_buffer_index, u8 priority)
 {
 	struct ivpu_device *vdev = file_priv->vdev;
 	struct ivpu_job *job;
@@ -812,7 +845,8 @@ static int ivpu_submit(struct drm_file *file, struct ivpu_file_priv *file_priv,
 		goto err_exit_dev;
 	}
 
-	ret = ivpu_job_prepare_bos_for_submit(file, job, buf_handles, buffer_count, cmds_offset);
+	ret = ivpu_job_prepare_bos_for_submit(file, job, buf_handles, buffer_count, cmds_offset,
+					      preempt_buffer_index);
 	if (ret) {
 		ivpu_err(vdev, "Failed to prepare job: %d\n", ret);
 		goto err_destroy_job;
@@ -866,7 +900,7 @@ int ivpu_submit_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
 	priority = ivpu_job_to_jsm_priority(args->priority);
 
 	return ivpu_submit(file, file_priv, 0, args->buffer_count, args->engine,
-			   (void __user *)args->buffers_ptr, args->commands_offset, priority);
+			   (void __user *)args->buffers_ptr, args->commands_offset, 0, priority);
 }
 
 int ivpu_cmdq_submit_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
@@ -883,6 +917,9 @@ int ivpu_cmdq_submit_ioctl(struct drm_device *dev, void *data, struct drm_file *
 	if (args->buffer_count == 0 || args->buffer_count > JOB_MAX_BUFFER_COUNT)
 		return -EINVAL;
 
+	if (args->preempt_buffer_index >= args->buffer_count)
+		return -EINVAL;
+
 	if (!IS_ALIGNED(args->commands_offset, 8))
 		return -EINVAL;
 
@@ -893,7 +930,8 @@ int ivpu_cmdq_submit_ioctl(struct drm_device *dev, void *data, struct drm_file *
 		return -EBADFD;
 
 	return ivpu_submit(file, file_priv, args->cmdq_id, args->buffer_count, VPU_ENGINE_COMPUTE,
-			   (void __user *)args->buffers_ptr, args->commands_offset, 0);
+			   (void __user *)args->buffers_ptr, args->commands_offset,
+			   args->preempt_buffer_index, 0);
 }
 
 int ivpu_cmdq_create_ioctl(struct drm_device *dev, void *data, struct drm_file *file)
@@ -1012,7 +1050,7 @@ void ivpu_context_abort_work_fn(struct work_struct *work)
 
 	if (vdev->fw->sched_mode == VPU_SCHEDULING_MODE_HW)
 		if (ivpu_jsm_reset_engine(vdev, 0))
-			return;
+			goto runtime_put;
 
 	mutex_lock(&vdev->context_list_lock);
 	xa_for_each(&vdev->context_xa, ctx_id, file_priv) {
@@ -1036,7 +1074,7 @@ void ivpu_context_abort_work_fn(struct work_struct *work)
 		goto runtime_put;
 
 	if (ivpu_jsm_hws_resume_engine(vdev, 0))
-		return;
+		goto runtime_put;
 	/*
 	 * In hardware scheduling mode NPU already has stopped processing jobs
 	 * and won't send us any further notifications, thus we have to free job related resources
diff --git a/drivers/accel/ivpu/ivpu_job.h b/drivers/accel/ivpu/ivpu_job.h
index 2e301c2eea7b..d2fc4c151614 100644
--- a/drivers/accel/ivpu/ivpu_job.h
+++ b/drivers/accel/ivpu/ivpu_job.h
@@ -1,6 +1,6 @@
 /* SPDX-License-Identifier: GPL-2.0-only */
 /*
- * Copyright (C) 2020-2024 Intel Corporation
+ * Copyright (C) 2020-2025 Intel Corporation
  */
 
 #ifndef __IVPU_JOB_H__
@@ -15,12 +15,17 @@ struct ivpu_device;
 struct ivpu_file_priv;
 
 /**
- * struct ivpu_cmdq - Object representing device queue used to send jobs.
- * @jobq:	   Pointer to job queue memory shared with the device
- * @mem:           Memory allocated for the job queue, shared with device
- * @entry_count    Number of job entries in the queue
- * @db_id:	   Doorbell assigned to this job queue
- * @db_registered: True if doorbell is registered in device
+ * struct ivpu_cmdq - Represents a command queue for submitting jobs to the VPU.
+ * Tracks queue memory, preemption buffers, and metadata for job management.
+ * @jobq:                Pointer to job queue memory shared with the device
+ * @primary_preempt_buf: Primary preemption buffer for this queue (optional)
+ * @secondary_preempt_buf: Secondary preemption buffer for this queue (optional)
+ * @mem:                 Memory allocated for the job queue, shared with device
+ * @entry_count:         Number of job entries in the queue
+ * @id:                  Unique command queue ID
+ * @db_id:               Doorbell ID assigned to this job queue
+ * @priority:            Priority level of the command queue
+ * @is_legacy:           True if this is a legacy command queue
  */
 struct ivpu_cmdq {
 	struct vpu_job_queue *jobq;
@@ -35,16 +40,21 @@ struct ivpu_cmdq {
 };
 
 /**
- * struct ivpu_job - KMD object that represents batchbuffer / DMA buffer.
- * Each batch / DMA buffer is a job to be submitted and executed by the VPU FW.
- * This is a unit of execution, and be tracked by the job_id for
- * any status reporting from VPU FW through IPC JOB RET/DONE message.
- * @file_priv:		  The client that submitted this job
- * @job_id:		  Job ID for KMD tracking and job status reporting from VPU FW
- * @status:		  Status of the Job from IPC JOB RET/DONE message
- * @batch_buffer:	  CPU vaddr points to the batch buffer memory allocated for the job
- * @submit_status_offset: Offset within batch buffer where job completion handler
-			  will update the job status
+ * struct ivpu_job - Representing a batch or DMA buffer submitted to the VPU.
+ * Each job is a unit of execution, tracked by job_id for status reporting from VPU FW.
+ * The structure holds all resources and metadata needed for job submission, execution,
+ * and completion handling.
+ * @vdev:                Pointer to the VPU device
+ * @file_priv:           The client context that submitted this job
+ * @done_fence:          Fence signaled when job completes
+ * @cmd_buf_vpu_addr:    VPU address of the command buffer for this job
+ * @cmdq_id:             Command queue ID used for submission
+ * @job_id:              Unique job ID for tracking and status reporting
+ * @engine_idx:          Engine index for job execution
+ * @primary_preempt_buf: Primary preemption buffer for job
+ * @secondary_preempt_buf: Secondary preemption buffer for job (optional)
+ * @bo_count:            Number of buffer objects associated with this job
+ * @bos:                 Array of buffer objects used by the job (batch buffer is at index 0)
  */
 struct ivpu_job {
 	struct ivpu_device *vdev;
@@ -54,6 +64,8 @@ struct ivpu_job {
 	u32 cmdq_id;
 	u32 job_id;
 	u32 engine_idx;
+	struct ivpu_bo *primary_preempt_buf;
+	struct ivpu_bo *secondary_preempt_buf;
 	size_t bo_count;
 	struct ivpu_bo *bos[] __counted_by(bo_count);
 };
diff --git a/drivers/accel/ivpu/ivpu_mmu_context.c b/drivers/accel/ivpu/ivpu_mmu_context.c
index f0267efa55aa..4ffc783426be 100644
--- a/drivers/accel/ivpu/ivpu_mmu_context.c
+++ b/drivers/accel/ivpu/ivpu_mmu_context.c
@@ -568,7 +568,7 @@ void ivpu_mmu_context_init(struct ivpu_device *vdev, struct ivpu_mmu_context *ct
 	mutex_init(&ctx->lock);
 
 	if (!context_id) {
-		start = vdev->hw->ranges.global.start;
+		start = vdev->hw->ranges.runtime.start;
 		end = vdev->hw->ranges.shave.end;
 	} else {
 		start = min_t(u64, vdev->hw->ranges.user.start, vdev->hw->ranges.shave.start);
diff --git a/drivers/accel/ivpu/ivpu_pm.c b/drivers/accel/ivpu/ivpu_pm.c
index 475ddc94f1cf..63c95307faa1 100644
--- a/drivers/accel/ivpu/ivpu_pm.c
+++ b/drivers/accel/ivpu/ivpu_pm.c
@@ -54,7 +54,7 @@ static void ivpu_pm_prepare_cold_boot(struct ivpu_device *vdev)
 static void ivpu_pm_prepare_warm_boot(struct ivpu_device *vdev)
 {
 	struct ivpu_fw_info *fw = vdev->fw;
-	struct vpu_boot_params *bp = ivpu_bo_vaddr(fw->mem);
+	struct vpu_boot_params *bp = ivpu_bo_vaddr(fw->mem_bp);
 
 	if (!bp->save_restore_ret_address) {
 		ivpu_pm_prepare_cold_boot(vdev);
@@ -502,6 +502,11 @@ void ivpu_pm_irq_dct_work_fn(struct work_struct *work)
 	else
 		ret = ivpu_pm_dct_disable(vdev);
 
-	if (!ret)
-		ivpu_hw_btrs_dct_set_status(vdev, enable, vdev->pm->dct_active_percent);
+	if (!ret) {
+		/* Convert percent to U1.7 format */
+		u8 val = DIV_ROUND_CLOSEST(vdev->pm->dct_active_percent * 128, 100);
+
+		ivpu_hw_btrs_dct_set_status(vdev, enable, val);
+	}
+
 }
diff --git a/drivers/accel/ivpu/vpu_jsm_api.h b/drivers/accel/ivpu/vpu_jsm_api.h
index 4b6b2b3d2583..de1b37ea1251 100644
--- a/drivers/accel/ivpu/vpu_jsm_api.h
+++ b/drivers/accel/ivpu/vpu_jsm_api.h
@@ -1,15 +1,16 @@
 /* SPDX-License-Identifier: MIT */
 /*
- * Copyright (c) 2020-2024, Intel Corporation.
+ * Copyright (c) 2020-2025, Intel Corporation.
+ */
+
+/**
+ * @addtogroup Jsm
+ * @{
  */
 
 /**
  * @file
  * @brief JSM shared definitions
- *
- * @ingroup Jsm
- * @brief JSM shared definitions
- * @{
  */
 #ifndef VPU_JSM_API_H
 #define VPU_JSM_API_H
@@ -22,12 +23,12 @@
 /*
  * Minor version changes when API backward compatibility is preserved.
  */
-#define VPU_JSM_API_VER_MINOR 29
+#define VPU_JSM_API_VER_MINOR 32
 
 /*
  * API header changed (field names, documentation, formatting) but API itself has not been changed
  */
-#define VPU_JSM_API_VER_PATCH 0
+#define VPU_JSM_API_VER_PATCH 5
 
 /*
  * Index in the API version table
@@ -71,9 +72,12 @@
 #define VPU_JSM_STATUS_MVNCI_OUT_OF_RESOURCES		 0xAU
 #define VPU_JSM_STATUS_MVNCI_NOT_IMPLEMENTED		 0xBU
 #define VPU_JSM_STATUS_MVNCI_INTERNAL_ERROR		 0xCU
-/* Job status returned when the job was preempted mid-inference */
+/* @deprecated (use VPU_JSM_STATUS_PREEMPTED_MID_COMMAND instead) */
 #define VPU_JSM_STATUS_PREEMPTED_MID_INFERENCE		 0xDU
+/* Job status returned when the job was preempted mid-command */
+#define VPU_JSM_STATUS_PREEMPTED_MID_COMMAND		 0xDU
 #define VPU_JSM_STATUS_MVNCI_CONTEXT_VIOLATION_HW	 0xEU
+#define VPU_JSM_STATUS_MVNCI_PREEMPTION_TIMED_OUT	 0xFU
 
 /*
  * Host <-> VPU IPC channels.
@@ -134,11 +138,21 @@ enum {
 	 *  2. Native fence queues are only supported on VPU 40xx onwards.
 	 */
 	VPU_JOB_QUEUE_FLAGS_USE_NATIVE_FENCE_MASK = (1 << 1U),
-
 	/*
 	 * Enable turbo mode for testing NPU performance; not recommended for regular usage.
 	 */
-	VPU_JOB_QUEUE_FLAGS_TURBO_MODE = (1 << 2U)
+	VPU_JOB_QUEUE_FLAGS_TURBO_MODE = (1 << 2U),
+	/*
+	 * Queue error detection mode flag
+	 * For 'interactive' queues (this bit not set), the FW will identify queues that have not
+	 * completed a job inside the TDR timeout as in error as part of engine reset sequence.
+	 * For 'non-interactive' queues (this bit set), the FW will identify queues that have not
+	 * progressed the heartbeat inside the non-interactive no-progress timeout as in error as
+	 * part of engine reset sequence. Additionally, there is an upper limit applied to these
+	 * queues: even if they progress the heartbeat, if they run longer than non-interactive
+	 * timeout, then the FW will also identify them as in error.
+	 */
+	VPU_JOB_QUEUE_FLAGS_NON_INTERACTIVE = (1 << 3U)
 };
 
 /*
@@ -209,7 +223,7 @@ enum {
  */
 #define VPU_INLINE_CMD_TYPE_FENCE_SIGNAL 0x2
 
-/*
+/**
  * Job scheduling priority bands for both hardware scheduling and OS scheduling.
  */
 enum vpu_job_scheduling_priority_band {
@@ -220,16 +234,16 @@ enum vpu_job_scheduling_priority_band {
 	VPU_JOB_SCHEDULING_PRIORITY_BAND_COUNT = 4,
 };
 
-/*
+/**
  * Job format.
  * Jobs defines the actual workloads to be executed by a given engine.
  */
 struct vpu_job_queue_entry {
-	/**< Address of VPU commands batch buffer */
+	/** Address of VPU commands batch buffer */
 	u64 batch_buf_addr;
-	/**< Job ID */
+	/** Job ID */
 	u32 job_id;
-	/**< Flags bit field, see VPU_JOB_FLAGS_* above */
+	/** Flags bit field, see VPU_JOB_FLAGS_* above */
 	u32 flags;
 	/**
 	 * Doorbell ring timestamp taken by KMD from SoC's global system clock, in
@@ -237,20 +251,20 @@ struct vpu_job_queue_entry {
 	 * to match other profiling timestamps.
 	 */
 	u64 doorbell_timestamp;
-	/**< Extra id for job tracking, used only in the firmware perf traces */
+	/** Extra id for job tracking, used only in the firmware perf traces */
 	u64 host_tracking_id;
-	/**< Address of the primary preemption buffer to use for this job */
+	/** Address of the primary preemption buffer to use for this job */
 	u64 primary_preempt_buf_addr;
-	/**< Size of the primary preemption buffer to use for this job */
+	/** Size of the primary preemption buffer to use for this job */
 	u32 primary_preempt_buf_size;
-	/**< Size of secondary preemption buffer to use for this job */
+	/** Size of secondary preemption buffer to use for this job */
 	u32 secondary_preempt_buf_size;
-	/**< Address of secondary preemption buffer to use for this job */
+	/** Address of secondary preemption buffer to use for this job */
 	u64 secondary_preempt_buf_addr;
 	u64 reserved_0;
 };
 
-/*
+/**
  * Inline command format.
  * Inline commands are the commands executed at scheduler level (typically,
  * synchronization directives). Inline command and job objects must be of
@@ -258,34 +272,36 @@ struct vpu_job_queue_entry {
  */
 struct vpu_inline_cmd {
 	u64 reserved_0;
-	/* Inline command type, see VPU_INLINE_CMD_TYPE_* defines. */
+	/** Inline command type, see VPU_INLINE_CMD_TYPE_* defines. */
 	u32 type;
-	/* Flags bit field, see VPU_JOB_FLAGS_* above. */
+	/** Flags bit field, see VPU_JOB_FLAGS_* above. */
 	u32 flags;
-	/* Inline command payload. Depends on inline command type. */
-	union {
-		/* Fence (wait and signal) commands' payload. */
-		struct {
-			/* Fence object handle. */
+	/** Inline command payload. Depends on inline command type. */
+	union payload {
+		/** Fence (wait and signal) commands' payload. */
+		struct fence {
+			/** Fence object handle. */
 			u64 fence_handle;
-			/* User VA of the current fence value. */
+			/** User VA of the current fence value. */
 			u64 current_value_va;
-			/* User VA of the monitored fence value (read-only). */
+			/** User VA of the monitored fence value (read-only). */
 			u64 monitored_value_va;
-			/* Value to wait for or write in fence location. */
+			/** Value to wait for or write in fence location. */
 			u64 value;
-			/* User VA of the log buffer in which to add log entry on completion. */
+			/** User VA of the log buffer in which to add log entry on completion. */
 			u64 log_buffer_va;
-			/* NPU private data. */
+			/** NPU private data. */
 			u64 npu_private_data;
 		} fence;
-		/* Other commands do not have a payload. */
-		/* Payload definition for future inline commands can be inserted here. */
+		/**
+		 * Other commands do not have a payload:
+		 * Payload definition for future inline commands can be inserted here.
+		 */
 		u64 reserved_1[6];
 	} payload;
 };
 
-/*
+/**
  * Job queue slots can be populated either with job objects or inline command objects.
  */
 union vpu_jobq_slot {
@@ -293,7 +309,7 @@ union vpu_jobq_slot {
 	struct vpu_inline_cmd inline_cmd;
 };
 
-/*
+/**
  * Job queue control registers.
  */
 struct vpu_job_queue_header {
@@ -301,18 +317,18 @@ struct vpu_job_queue_header {
 	u32 head;
 	u32 tail;
 	u32 flags;
-	/* Set to 1 to indicate priority_band field is valid */
+	/** Set to 1 to indicate priority_band field is valid */
 	u32 priority_band_valid;
-	/*
+	/**
 	 * Priority for the work of this job queue, valid only if the HWS is NOT used
-	 * and the `priority_band_valid` is set to 1. It is applied only during
-	 * the VPU_JSM_MSG_REGISTER_DB message processing.
-	 * The device firmware might use the `priority_band` to optimize the power
+	 * and the @ref priority_band_valid is set to 1. It is applied only during
+	 * the @ref VPU_JSM_MSG_REGISTER_DB message processing.
+	 * The device firmware might use the priority_band to optimize the power
 	 * management logic, but it will not affect the order of jobs.
 	 * Available priority bands: @see enum vpu_job_scheduling_priority_band
 	 */
 	u32 priority_band;
-	/* Inside realtime band assigns a further priority, limited to 0..31 range */
+	/** Inside realtime band assigns a further priority, limited to 0..31 range */
 	u32 realtime_priority_level;
 	u32 reserved_0[9];
 };
@@ -337,16 +353,16 @@ enum vpu_trace_entity_type {
 	VPU_TRACE_ENTITY_TYPE_HW_COMPONENT = 2,
 };
 
-/*
+/**
  * HWS specific log buffer header details.
  * Total size is 32 bytes.
  */
 struct vpu_hws_log_buffer_header {
-	/* Written by VPU after adding a log entry. Initialised by host to 0. */
+	/** Written by VPU after adding a log entry. Initialised by host to 0. */
 	u32 first_free_entry_index;
-	/* Incremented by VPU every time the VPU writes the 0th entry; initialised by host to 0. */
+	/** Incremented by VPU every time the VPU writes the 0th entry; initialised by host to 0. */
 	u32 wraparound_count;
-	/*
+	/**
 	 * This is the number of buffers that can be stored in the log buffer provided by the host.
 	 * It is written by host before passing buffer to VPU. VPU should consider it read-only.
 	 */
@@ -354,14 +370,14 @@ struct vpu_hws_log_buffer_header {
 	u64 reserved[2];
 };
 
-/*
+/**
  * HWS specific log buffer entry details.
  * Total size is 32 bytes.
  */
 struct vpu_hws_log_buffer_entry {
-	/* VPU timestamp must be an invariant timer tick (not impacted by DVFS) */
+	/** VPU timestamp must be an invariant timer tick (not impacted by DVFS) */
 	u64 vpu_timestamp;
-	/*
+	/**
 	 * Operation type:
 	 *     0 - context state change
 	 *     1 - queue new work
@@ -371,7 +387,7 @@ struct vpu_hws_log_buffer_entry {
 	 */
 	u32 operation_type;
 	u32 reserved;
-	/* Operation data depends on operation type */
+	/** Operation data depends on operation type */
 	u64 operation_data[2];
 };
 
@@ -381,51 +397,54 @@ enum vpu_hws_native_fence_log_type {
 	VPU_HWS_NATIVE_FENCE_LOG_TYPE_SIGNALS = 2
 };
 
-/* HWS native fence log buffer header. */
+/** HWS native fence log buffer header. */
 struct vpu_hws_native_fence_log_header {
 	union {
 		struct {
-			/* Index of the first free entry in buffer. */
+			/** Index of the first free entry in buffer. */
 			u32 first_free_entry_idx;
-			/* Incremented each time NPU wraps around the buffer to write next entry. */
+			/**
+			 * Incremented each time NPU wraps around
+			 * the buffer to write next entry.
+			 */
 			u32 wraparound_count;
 		};
-		/* Field allowing atomic update of both fields above. */
+		/** Field allowing atomic update of both fields above. */
 		u64 atomic_wraparound_and_entry_idx;
 	};
-	/* Log buffer type, see enum vpu_hws_native_fence_log_type. */
+	/** Log buffer type, see enum vpu_hws_native_fence_log_type. */
 	u64 type;
-	/* Allocated number of entries in the log buffer. */
+	/** Allocated number of entries in the log buffer. */
 	u64 entry_nb;
 	u64 reserved[2];
 };
 
-/* Native fence log operation types. */
+/** Native fence log operation types. */
 enum vpu_hws_native_fence_log_op {
 	VPU_HWS_NATIVE_FENCE_LOG_OP_SIGNAL_EXECUTED = 0,
 	VPU_HWS_NATIVE_FENCE_LOG_OP_WAIT_UNBLOCKED = 1
 };
 
-/* HWS native fence log entry. */
+/** HWS native fence log entry. */
 struct vpu_hws_native_fence_log_entry {
-	/* Newly signaled/unblocked fence value. */
+	/** Newly signaled/unblocked fence value. */
 	u64 fence_value;
-	/* Native fence object handle to which this operation belongs. */
+	/** Native fence object handle to which this operation belongs. */
 	u64 fence_handle;
-	/* Operation type, see enum vpu_hws_native_fence_log_op. */
+	/** Operation type, see enum vpu_hws_native_fence_log_op. */
 	u64 op_type;
 	u64 reserved_0;
-	/*
+	/**
 	 * VPU_HWS_NATIVE_FENCE_LOG_OP_WAIT_UNBLOCKED only: Timestamp at which fence
 	 * wait was started (in NPU SysTime).
 	 */
 	u64 fence_wait_start_ts;
 	u64 reserved_1;
-	/* Timestamp at which fence operation was completed (in NPU SysTime). */
+	/** Timestamp at which fence operation was completed (in NPU SysTime). */
 	u64 fence_end_ts;
 };
 
-/* Native fence log buffer. */
+/** Native fence log buffer. */
 struct vpu_hws_native_fence_log_buffer {
 	struct vpu_hws_native_fence_log_header header;
 	struct vpu_hws_native_fence_log_entry entry[];
@@ -450,8 +469,21 @@ enum vpu_ipc_msg_type {
 	 * after preemption or when resubmitting jobs to the queue.
 	 */
 	VPU_JSM_MSG_ENGINE_PREEMPT = 0x1101,
+	/**
+	 * OS scheduling doorbell register command
+	 * @see vpu_ipc_msg_payload_register_db
+	 */
 	VPU_JSM_MSG_REGISTER_DB = 0x1102,
+	/**
+	 * OS scheduling doorbell unregister command
+	 * @see vpu_ipc_msg_payload_unregister_db
+	 */
 	VPU_JSM_MSG_UNREGISTER_DB = 0x1103,
+	/**
+	 * Query engine heartbeat. Heartbeat is expected to increase monotonically
+	 * and increase while work is being progressed by NPU.
+	 * @see vpu_ipc_msg_payload_query_engine_hb
+	 */
 	VPU_JSM_MSG_QUERY_ENGINE_HB = 0x1104,
 	VPU_JSM_MSG_GET_POWER_LEVEL_COUNT = 0x1105,
 	VPU_JSM_MSG_GET_POWER_LEVEL = 0x1106,
@@ -477,6 +509,7 @@ enum vpu_ipc_msg_type {
 	 * aborted and removed from internal scheduling queues. All doorbells assigned
 	 * to the host_ssid are unregistered and any internal FW resources belonging to
 	 * the host_ssid are released.
+	 * @see vpu_ipc_msg_payload_ssid_release
 	 */
 	VPU_JSM_MSG_SSID_RELEASE = 0x110e,
 	/**
@@ -504,26 +537,51 @@ enum vpu_ipc_msg_type {
 	 * @see vpu_jsm_metric_streamer_start
 	 */
 	VPU_JSM_MSG_METRIC_STREAMER_INFO = 0x1112,
-	/** Control command: Priority band setup */
+	/**
+	 * Control command: Priority band setup
+	 * @see vpu_ipc_msg_payload_hws_priority_band_setup
+	 */
 	VPU_JSM_MSG_SET_PRIORITY_BAND_SETUP = 0x1113,
-	/** Control command: Create command queue */
+	/**
+	 * Control command: Create command queue
+	 * @see vpu_ipc_msg_payload_hws_create_cmdq
+	 */
 	VPU_JSM_MSG_CREATE_CMD_QUEUE = 0x1114,
-	/** Control command: Destroy command queue */
+	/**
+	 * Control command: Destroy command queue
+	 * @see vpu_ipc_msg_payload_hws_destroy_cmdq
+	 */
 	VPU_JSM_MSG_DESTROY_CMD_QUEUE = 0x1115,
-	/** Control command: Set context scheduling properties */
+	/**
+	 * Control command: Set context scheduling properties
+	 * @see vpu_ipc_msg_payload_hws_set_context_sched_properties
+	 */
 	VPU_JSM_MSG_SET_CONTEXT_SCHED_PROPERTIES = 0x1116,
-	/*
+	/**
 	 * Register a doorbell to notify VPU of new work. The doorbell may later be
 	 * deallocated or reassigned to another context.
+	 * @see vpu_jsm_hws_register_db
 	 */
 	VPU_JSM_MSG_HWS_REGISTER_DB = 0x1117,
-	/** Control command: Log buffer setting */
+	/**
+	 * Control command: Log buffer setting
+	 * @see vpu_ipc_msg_payload_hws_set_scheduling_log
+	 */
 	VPU_JSM_MSG_HWS_SET_SCHEDULING_LOG = 0x1118,
-	/* Control command: Suspend command queue. */
+	/**
+	 * Control command: Suspend command queue.
+	 * @see vpu_ipc_msg_payload_hws_suspend_cmdq
+	 */
 	VPU_JSM_MSG_HWS_SUSPEND_CMDQ = 0x1119,
-	/* Control command: Resume command queue */
+	/**
+	 * Control command: Resume command queue
+	 * @see vpu_ipc_msg_payload_hws_resume_cmdq
+	 */
 	VPU_JSM_MSG_HWS_RESUME_CMDQ = 0x111a,
-	/* Control command: Resume engine after reset */
+	/**
+	 * Control command: Resume engine after reset
+	 * @see vpu_ipc_msg_payload_hws_resume_engine
+	 */
 	VPU_JSM_MSG_HWS_ENGINE_RESUME = 0x111b,
 	/* Control command: Enable survivability/DCT mode */
 	VPU_JSM_MSG_DCT_ENABLE = 0x111c,
@@ -540,7 +598,8 @@ enum vpu_ipc_msg_type {
 	VPU_JSM_MSG_BLOB_DEINIT_DEPRECATED = VPU_JSM_MSG_GENERAL_CMD,
 	/**
 	 * Control dyndbg behavior by executing a dyndbg command; equivalent to
-	 * Linux command: `echo '<dyndbg_cmd>' > <debugfs>/dynamic_debug/control`.
+	 * Linux command:
+	 * @verbatim echo '<dyndbg_cmd>' > <debugfs>/dynamic_debug/control @endverbatim
 	 */
 	VPU_JSM_MSG_DYNDBG_CONTROL = 0x1201,
 	/**
@@ -550,15 +609,26 @@ enum vpu_ipc_msg_type {
 
 	/* IPC Device -> Host, Job completion */
 	VPU_JSM_MSG_JOB_DONE = 0x2100,
-	/* IPC Device -> Host, Fence signalled */
+	/**
+	 * IPC Device -> Host, Fence signalled
+	 * @see vpu_ipc_msg_payload_native_fence_signalled
+	 */
 	VPU_JSM_MSG_NATIVE_FENCE_SIGNALLED = 0x2101,
 
 	/* IPC Device -> Host, Async command completion */
 	VPU_JSM_MSG_ASYNC_CMD_DONE = 0x2200,
+	/**
+	 * IPC Device -> Host, engine reset complete
+	 * @see vpu_ipc_msg_payload_engine_reset_done
+	 */
 	VPU_JSM_MSG_ENGINE_RESET_DONE = VPU_JSM_MSG_ASYNC_CMD_DONE,
 	VPU_JSM_MSG_ENGINE_PREEMPT_DONE = 0x2201,
 	VPU_JSM_MSG_REGISTER_DB_DONE = 0x2202,
 	VPU_JSM_MSG_UNREGISTER_DB_DONE = 0x2203,
+	/**
+	 * Response to query engine heartbeat.
+	 * @see vpu_ipc_msg_payload_query_engine_hb_done
+	 */
 	VPU_JSM_MSG_QUERY_ENGINE_HB_DONE = 0x2204,
 	VPU_JSM_MSG_GET_POWER_LEVEL_COUNT_DONE = 0x2205,
 	VPU_JSM_MSG_GET_POWER_LEVEL_DONE = 0x2206,
@@ -575,7 +645,10 @@ enum vpu_ipc_msg_type {
 	VPU_JSM_MSG_TRACE_GET_CAPABILITY_RSP = 0x220c,
 	/** Response to VPU_JSM_MSG_TRACE_GET_NAME. */
 	VPU_JSM_MSG_TRACE_GET_NAME_RSP = 0x220d,
-	/** Response to VPU_JSM_MSG_SSID_RELEASE. */
+	/**
+	 * Response to VPU_JSM_MSG_SSID_RELEASE.
+	 * @see vpu_ipc_msg_payload_ssid_release
+	 */
 	VPU_JSM_MSG_SSID_RELEASE_DONE = 0x220e,
 	/**
 	 * Response to VPU_JSM_MSG_METRIC_STREAMER_START.
@@ -605,29 +678,56 @@ enum vpu_ipc_msg_type {
 	/**
 	 * Asynchronous event sent from the VPU to the host either when the current
 	 * metric buffer is full or when the VPU has collected a multiple of
-	 * @notify_sample_count samples as indicated through the start command
-	 * (VPU_JSM_MSG_METRIC_STREAMER_START). Returns information about collected
-	 * metric data.
+	 * @ref vpu_jsm_metric_streamer_start::notify_sample_count samples as indicated
+	 * through the start command (VPU_JSM_MSG_METRIC_STREAMER_START). Returns
+	 * information about collected metric data.
 	 * @see vpu_jsm_metric_streamer_done
 	 */
 	VPU_JSM_MSG_METRIC_STREAMER_NOTIFICATION = 0x2213,
-	/** Response to control command: Priority band setup */
+	/**
+	 * Response to control command: Priority band setup
+	 * @see vpu_ipc_msg_payload_hws_priority_band_setup
+	 */
 	VPU_JSM_MSG_SET_PRIORITY_BAND_SETUP_RSP = 0x2214,
-	/** Response to control command: Create command queue */
+	/**
+	 * Response to control command: Create command queue
+	 * @see vpu_ipc_msg_payload_hws_create_cmdq_rsp
+	 */
 	VPU_JSM_MSG_CREATE_CMD_QUEUE_RSP = 0x2215,
-	/** Response to control command: Destroy command queue */
+	/**
+	 * Response to control command: Destroy command queue
+	 * @see vpu_ipc_msg_payload_hws_destroy_cmdq
+	 */
 	VPU_JSM_MSG_DESTROY_CMD_QUEUE_RSP = 0x2216,
-	/** Response to control command: Set context scheduling properties */
+	/**
+	 * Response to control command: Set context scheduling properties
+	 * @see vpu_ipc_msg_payload_hws_set_context_sched_properties
+	 */
 	VPU_JSM_MSG_SET_CONTEXT_SCHED_PROPERTIES_RSP = 0x2217,
-	/** Response to control command: Log buffer setting */
+	/**
+	 * Response to control command: Log buffer setting
+	 * @see vpu_ipc_msg_payload_hws_set_scheduling_log
+	 */
 	VPU_JSM_MSG_HWS_SET_SCHEDULING_LOG_RSP = 0x2218,
-	/* IPC Device -> Host, HWS notify index entry of log buffer written */
+	/**
+	 * IPC Device -> Host, HWS notify index entry of log buffer written
+	 * @see vpu_ipc_msg_payload_hws_scheduling_log_notification
+	 */
 	VPU_JSM_MSG_HWS_SCHEDULING_LOG_NOTIFICATION = 0x2219,
-	/* IPC Device -> Host, HWS completion of a context suspend request */
+	/**
+	 * IPC Device -> Host, HWS completion of a context suspend request
+	 * @see vpu_ipc_msg_payload_hws_suspend_cmdq
+	 */
 	VPU_JSM_MSG_HWS_SUSPEND_CMDQ_DONE = 0x221a,
-	/* Response to control command: Resume command queue */
+	/**
+	 * Response to control command: Resume command queue
+	 * @see vpu_ipc_msg_payload_hws_resume_cmdq
+	 */
 	VPU_JSM_MSG_HWS_RESUME_CMDQ_RSP = 0x221b,
-	/* Response to control command: Resume engine command response */
+	/**
+	 * Response to control command: Resume engine command response
+	 * @see vpu_ipc_msg_payload_hws_resume_engine
+	 */
 	VPU_JSM_MSG_HWS_RESUME_ENGINE_DONE = 0x221c,
 	/* Response to control command: Enable survivability/DCT mode */
 	VPU_JSM_MSG_DCT_ENABLE_DONE = 0x221d,
@@ -670,40 +770,44 @@ struct vpu_ipc_msg_payload_engine_preempt {
 	u32 preempt_id;
 };
 
-/*
- * @brief Register doorbell command structure.
+/**
+ * Register doorbell command structure.
  * This structure supports doorbell registration for only OS scheduling.
  * @see VPU_JSM_MSG_REGISTER_DB
  */
 struct vpu_ipc_msg_payload_register_db {
-	/* Index of the doorbell to register. */
+	/** Index of the doorbell to register. */
 	u32 db_idx;
-	/* Reserved */
+	/** Reserved */
 	u32 reserved_0;
-	/* Virtual address in Global GTT pointing to the start of job queue. */
+	/** Virtual address in Global GTT pointing to the start of job queue. */
 	u64 jobq_base;
-	/* Size of the job queue in bytes. */
+	/** Size of the job queue in bytes. */
 	u32 jobq_size;
-	/* Host sub-stream ID for the context assigned to the doorbell. */
+	/** Host sub-stream ID for the context assigned to the doorbell. */
 	u32 host_ssid;
 };
 
 /**
- * @brief Unregister doorbell command structure.
+ * Unregister doorbell command structure.
  * Request structure to unregister a doorbell for both HW and OS scheduling.
  * @see VPU_JSM_MSG_UNREGISTER_DB
  */
 struct vpu_ipc_msg_payload_unregister_db {
-	/* Index of the doorbell to unregister. */
+	/** Index of the doorbell to unregister. */
 	u32 db_idx;
-	/* Reserved */
+	/** Reserved */
 	u32 reserved_0;
 };
 
+/**
+ * Heartbeat request structure
+ * @see VPU_JSM_MSG_QUERY_ENGINE_HB
+ */
 struct vpu_ipc_msg_payload_query_engine_hb {
-	/* Engine to return heartbeat value. */
+	/** Engine to return heartbeat value. */
 	u32 engine_idx;
-	/* Reserved */
+	/** Reserved */
 	u32 reserved_0;
 };
 
@@ -723,10 +827,14 @@ struct vpu_ipc_msg_payload_power_level {
 	u32 reserved_0;
 };
 
+/**
+ * Structure for requesting ssid release
+ * @see VPU_JSM_MSG_SSID_RELEASE
+ */
 struct vpu_ipc_msg_payload_ssid_release {
-	/* Host sub-stream ID for the context to be released. */
+	/** Host sub-stream ID for the context to be released. */
 	u32 host_ssid;
-	/* Reserved */
+	/** Reserved */
 	u32 reserved_0;
 };
 
@@ -752,7 +860,7 @@ struct vpu_jsm_metric_streamer_start {
 	u64 sampling_rate;
 	/**
 	 * If > 0 the VPU will send a VPU_JSM_MSG_METRIC_STREAMER_NOTIFICATION message
-	 * after every @notify_sample_count samples is collected or dropped by the VPU.
+	 * after every @ref notify_sample_count samples is collected or dropped by the VPU.
 	 * If set to UINT_MAX the VPU will only generate a notification when the metric
 	 * buffer is full. If set to 0 the VPU will never generate a notification.
 	 */
@@ -762,9 +870,9 @@ struct vpu_jsm_metric_streamer_start {
 	 * Address and size of the buffer where the VPU will write metric data. The
 	 * VPU writes all counters from enabled metric groups one after another. If
 	 * there is no space left to write data at the next sample period the VPU
-	 * will switch to the next buffer (@see next_buffer_addr) and will optionally
-	 * send a notification to the host driver if @notify_sample_count is non-zero.
-	 * If @next_buffer_addr is NULL the VPU will stop collecting metric data.
+	 * will switch to the next buffer (@ref next_buffer_addr) and will optionally
+	 * send a notification to the host driver if @ref notify_sample_count is non-zero.
+	 * If @ref next_buffer_addr is NULL the VPU will stop collecting metric data.
 	 */
 	u64 buffer_addr;
 	u64 buffer_size;
@@ -844,38 +952,47 @@ struct vpu_ipc_msg_payload_job_done {
 	u64 cmdq_id;
 };
 
-/*
+/**
  * Notification message upon native fence signalling.
  * @see VPU_JSM_MSG_NATIVE_FENCE_SIGNALLED
  */
 struct vpu_ipc_msg_payload_native_fence_signalled {
-	/* Engine ID. */
+	/** Engine ID. */
 	u32 engine_idx;
-	/* Host SSID. */
+	/** Host SSID. */
 	u32 host_ssid;
-	/* CMDQ ID */
+	/** CMDQ ID */
 	u64 cmdq_id;
-	/* Fence object handle. */
+	/** Fence object handle. */
 	u64 fence_handle;
 };
 
+/**
+ * vpu_ipc_msg_payload_engine_reset_done will contain an array of this structure
+ * which contains which queues caused reset if FW was able to detect any error.
+ * @see vpu_ipc_msg_payload_engine_reset_done
+ */
 struct vpu_jsm_engine_reset_context {
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved_0;
-	/* Command queue id */
+	/** Command queue id */
 	u64 cmdq_id;
-	/* See VPU_ENGINE_RESET_CONTEXT_* defines */
+	/** See VPU_ENGINE_RESET_CONTEXT_* defines */
 	u64 flags;
 };
 
+/**
+ * Engine reset response.
+ * @see VPU_JSM_MSG_ENGINE_RESET_DONE
+ */
 struct vpu_ipc_msg_payload_engine_reset_done {
-	/* Engine ordinal */
+	/** Engine ordinal */
 	u32 engine_idx;
-	/* Number of impacted contexts */
+	/** Number of impacted contexts */
 	u32 num_impacted_contexts;
-	/* Array of impacted command queue ids and their flags */
+	/** Array of impacted command queue ids and their flags */
 	struct vpu_jsm_engine_reset_context
 		impacted_contexts[VPU_MAX_ENGINE_RESET_IMPACTED_CONTEXTS];
 };
@@ -912,12 +1029,16 @@ struct vpu_ipc_msg_payload_unregister_db_done {
 	u32 reserved_0;
 };
 
+/**
+ * Structure for heartbeat response
+ * @see VPU_JSM_MSG_QUERY_ENGINE_HB_DONE
+ */
 struct vpu_ipc_msg_payload_query_engine_hb_done {
-	/* Engine returning heartbeat value. */
+	/** Engine returning heartbeat value. */
 	u32 engine_idx;
-	/* Reserved */
+	/** Reserved */
 	u32 reserved_0;
-	/* Heartbeat value. */
+	/** Heartbeat value. */
 	u64 heartbeat;
 };
 
@@ -937,7 +1058,10 @@ struct vpu_ipc_msg_payload_get_power_level_count_done {
 	u8 power_limit[16];
 };
 
-/* HWS priority band setup request / response */
+/**
+ * HWS priority band setup request / response
+ * @see VPU_JSM_MSG_SET_PRIORITY_BAND_SETUP
+ */
 struct vpu_ipc_msg_payload_hws_priority_band_setup {
 	/*
 	 * Grace period in 100ns units when preempting another priority band for
@@ -964,15 +1088,23 @@ struct vpu_ipc_msg_payload_hws_priority_band_setup {
 	 * TDR timeout value in milliseconds. Default value of 0 meaning no timeout.
 	 */
 	u32 tdr_timeout;
+	/* Non-interactive queue timeout for no progress of heartbeat in milliseconds.
+	 * Default value of 0 meaning no timeout.
+	 */
+	u32 non_interactive_no_progress_timeout;
+	/*
+	 * Non-interactive queue upper limit timeout value in milliseconds. Default
+	 * value of 0 meaning no timeout.
+	 */
+	u32 non_interactive_timeout;
 };
 
-/*
+/**
  * @brief HWS create command queue request.
  * Host will create a command queue via this command.
  * Note: Cmdq group is a handle of an object which
  * may contain one or more command queues.
  * @see VPU_JSM_MSG_CREATE_CMD_QUEUE
- * @see VPU_JSM_MSG_CREATE_CMD_QUEUE_RSP
  */
 struct vpu_ipc_msg_payload_hws_create_cmdq {
 	/* Process id */
@@ -993,66 +1125,73 @@ struct vpu_ipc_msg_payload_hws_create_cmdq {
 	u32 reserved_0;
 };
 
-/*
- * @brief HWS create command queue response.
- * @see VPU_JSM_MSG_CREATE_CMD_QUEUE
+/**
+ * HWS create command queue response.
  * @see VPU_JSM_MSG_CREATE_CMD_QUEUE_RSP
  */
 struct vpu_ipc_msg_payload_hws_create_cmdq_rsp {
-	/* Process id */
+	/** Process id */
 	u64 process_id;
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/* Engine for which queue is being created */
+	/** Engine for which queue is being created */
 	u32 engine_idx;
-	/* Command queue group */
+	/** Command queue group */
 	u64 cmdq_group;
-	/* Command queue id */
+	/** Command queue id */
 	u64 cmdq_id;
 };
 
-/* HWS destroy command queue request / response */
+/**
+ * HWS destroy command queue request / response
+ * @see VPU_JSM_MSG_DESTROY_CMD_QUEUE
+ * @see VPU_JSM_MSG_DESTROY_CMD_QUEUE_RSP
+ */
 struct vpu_ipc_msg_payload_hws_destroy_cmdq {
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved;
-	/* Command queue id */
+	/** Command queue id */
 	u64 cmdq_id;
 };
 
-/* HWS set context scheduling properties request / response */
+/**
+ * HWS set context scheduling properties request / response
+ * @see VPU_JSM_MSG_SET_CONTEXT_SCHED_PROPERTIES
+ * @see VPU_JSM_MSG_SET_CONTEXT_SCHED_PROPERTIES_RSP
+ */
 struct vpu_ipc_msg_payload_hws_set_context_sched_properties {
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved_0;
-	/* Command queue id */
+	/** Command queue id */
 	u64 cmdq_id;
-	/*
+	/**
 	 * Priority band to assign to work of this context.
 	 * Available priority bands: @see enum vpu_job_scheduling_priority_band
 	 */
 	u32 priority_band;
-	/* Inside realtime band assigns a further priority */
+	/** Inside realtime band assigns a further priority */
 	u32 realtime_priority_level;
-	/* Priority relative to other contexts in the same process */
+	/** Priority relative to other contexts in the same process */
 	s32 in_process_priority;
-	/* Zero padding / Reserved */
+	/** Zero padding / Reserved */
 	u32 reserved_1;
-	/*
+	/**
 	 * Context quantum relative to other contexts of same priority in the same process
 	 * Minimum value supported by NPU is 1ms (10000 in 100ns units).
 	 */
 	u64 context_quantum;
-	/* Grace period when preempting context of the same priority within the same process */
+	/** Grace period when preempting context of the same priority within the same process */
 	u64 grace_period_same_priority;
-	/* Grace period when preempting context of a lower priority within the same process */
+	/** Grace period when preempting context of a lower priority within the same process */
 	u64 grace_period_lower_priority;
 };
 
-/*
- * @brief Register doorbell command structure.
+/**
+ * Register doorbell command structure.
  * This structure supports doorbell registration for both HW and OS scheduling.
  * Note: Queue base and size are added here so that the same structure can be used for
  * OS scheduling and HW scheduling. For OS scheduling, cmdq_id will be ignored
@@ -1061,27 +1200,27 @@ struct vpu_ipc_msg_payload_hws_set_context_sched_properties {
  * @see VPU_JSM_MSG_HWS_REGISTER_DB
  */
 struct vpu_jsm_hws_register_db {
-	/* Index of the doorbell to register. */
+	/** Index of the doorbell to register. */
 	u32 db_id;
-	/* Host sub-stream ID for the context assigned to the doorbell. */
+	/** Host sub-stream ID for the context assigned to the doorbell. */
 	u32 host_ssid;
-	/* ID of the command queue associated with the doorbell. */
+	/** ID of the command queue associated with the doorbell. */
 	u64 cmdq_id;
-	/* Virtual address pointing to the start of command queue. */
+	/** Virtual address pointing to the start of command queue. */
 	u64 cmdq_base;
-	/* Size of the command queue in bytes. */
+	/** Size of the command queue in bytes. */
 	u64 cmdq_size;
 };
 
-/*
- * @brief Structure to set another buffer to be used for scheduling-related logging.
+/**
+ * Structure to set another buffer to be used for scheduling-related logging.
  * The size of the logging buffer and the number of entries is defined as part of the
  * buffer itself as described next.
  * The log buffer received from the host is made up of;
- *   - header:     32 bytes in size, as shown in 'struct vpu_hws_log_buffer_header'.
+ *   - header:     32 bytes in size, as shown in @ref vpu_hws_log_buffer_header.
  *                 The header contains the number of log entries in the buffer.
  *   - log entry:  0 to n-1, each log entry is 32 bytes in size, as shown in
- *                 'struct vpu_hws_log_buffer_entry'.
+ *                 @ref vpu_hws_log_buffer_entry.
  *                 The entry contains the VPU timestamp, operation type and data.
  * The host should provide the notify index value of log buffer to VPU. This is a
  * value defined within the log buffer and when written to will generate the
@@ -1095,30 +1234,30 @@ struct vpu_jsm_hws_register_db {
  * @see VPU_JSM_MSG_HWS_SCHEDULING_LOG_NOTIFICATION
  */
 struct vpu_ipc_msg_payload_hws_set_scheduling_log {
-	/* Engine ordinal */
+	/** Engine ordinal */
 	u32 engine_idx;
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/*
+	/**
 	 * VPU log buffer virtual address.
 	 * Set to 0 to disable logging for this engine.
 	 */
 	u64 vpu_log_buffer_va;
-	/*
+	/**
 	 * Notify index of log buffer. VPU_JSM_MSG_HWS_SCHEDULING_LOG_NOTIFICATION
 	 * is generated when an event log is written to this index.
 	 */
 	u64 notify_index;
-	/*
+	/**
 	 * Field is now deprecated, will be removed when KMD is updated to support removal
 	 */
 	u32 enable_extra_events;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved_0;
 };
 
-/*
- * @brief The scheduling log notification is generated by VPU when it writes
+/**
+ * The scheduling log notification is generated by VPU when it writes
  * an event into the log buffer at the notify_index. VPU notifies host with
  * VPU_JSM_MSG_HWS_SCHEDULING_LOG_NOTIFICATION. This is an asynchronous
  * message from VPU to host.
@@ -1126,14 +1265,14 @@ struct vpu_ipc_msg_payload_hws_set_scheduling_log {
  * @see VPU_JSM_MSG_HWS_SET_SCHEDULING_LOG
  */
 struct vpu_ipc_msg_payload_hws_scheduling_log_notification {
-	/* Engine ordinal */
+	/** Engine ordinal */
 	u32 engine_idx;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved_0;
 };
 
-/*
- * @brief HWS suspend command queue request and done structure.
+/**
+ * HWS suspend command queue request and done structure.
  * Host will request the suspend of contexts and VPU will;
  *   - Suspend all work on this context
  *   - Preempt any running work
@@ -1152,21 +1291,21 @@ struct vpu_ipc_msg_payload_hws_scheduling_log_notification {
  * @see VPU_JSM_MSG_HWS_SUSPEND_CMDQ_DONE
  */
 struct vpu_ipc_msg_payload_hws_suspend_cmdq {
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved_0;
-	/* Command queue id */
+	/** Command queue id */
 	u64 cmdq_id;
-	/*
+	/**
 	 * Suspend fence value - reported by the VPU suspend context
 	 * completed once suspend is complete.
 	 */
 	u64 suspend_fence_value;
 };
 
-/*
- * @brief HWS Resume command queue request / response structure.
+/**
+ * HWS Resume command queue request / response structure.
  * Host will request the resume of a context;
  *  - VPU will resume all work on this context
  *  - Scheduler will allow this context to be scheduled
@@ -1174,25 +1313,25 @@ struct vpu_ipc_msg_payload_hws_suspend_cmdq {
  * @see VPU_JSM_MSG_HWS_RESUME_CMDQ_RSP
  */
 struct vpu_ipc_msg_payload_hws_resume_cmdq {
-	/* Host SSID */
+	/** Host SSID */
 	u32 host_ssid;
-	/* Zero Padding */
+	/** Zero Padding */
 	u32 reserved_0;
-	/* Command queue id */
+	/** Command queue id */
 	u64 cmdq_id;
 };
 
-/*
- * @brief HWS Resume engine request / response structure.
- * After a HWS engine reset, all scheduling is stopped on VPU until a engine resume.
+/**
+ * HWS Resume engine request / response structure.
+ * After a HWS engine reset, all scheduling is stopped on VPU until an engine resume.
  * Host shall send this command to resume scheduling of any valid queue.
- * @see VPU_JSM_MSG_HWS_RESUME_ENGINE
+ * @see VPU_JSM_MSG_HWS_ENGINE_RESUME
  * @see VPU_JSM_MSG_HWS_RESUME_ENGINE_DONE
  */
 struct vpu_ipc_msg_payload_hws_resume_engine {
-	/* Engine to be resumed */
+	/** Engine to be resumed */
 	u32 engine_idx;
-	/* Reserved */
+	/** Reserved */
 	u32 reserved_0;
 };
 
@@ -1326,7 +1465,7 @@ struct vpu_jsm_metric_streamer_done {
 /**
  * Metric group description placed in the metric buffer after successful completion
  * of the VPU_JSM_MSG_METRIC_STREAMER_INFO command. This is followed by one or more
- * @vpu_jsm_metric_counter_descriptor records.
+ * @ref vpu_jsm_metric_counter_descriptor records.
  * @see VPU_JSM_MSG_METRIC_STREAMER_INFO
  */
 struct vpu_jsm_metric_group_descriptor {