Merge tag 'drm-next-2022-12-13' of git://anongit.freedesktop.org/drm/drm

Pull drm updates from Dave Airlie:
 "The biggest highlight is that the accel subsystem framework is merged.
  Hopefully for 6.3 we will be able to line up a driver to use it.

  In drivers land, i915 enables DG2 support by default now, and nouveau
  has a big stability refactoring and initial ampere support, AMD
  includes new hw IP support and should build on ARM again. There is
  also an ofdrm driver to take over offb on platforms it's used.

  Stuff outside my tree, the dma-buf patches hit a few places, the vc4
  firmware changes also do, and i915 has some interactions with MEI for
  discrete GPUs. I think all of those should have been acked/reviewed by
  relevant parties.

  New driver:
   - ofdrm - replacement for offb

  fbdev:
   - add support for nomodeset

  fourcc:
   - add Vivante tiled modifier

  core:
   - atomic-helpers: CRTC primary plane test fixes, fb access hooks
   - connector: TV API consistency, cmdline parser improvements
   - send connector hotplug on cleanup
   - sort makefile objects

  tests:
   - sort kunit tests
   - improve DP-MST tests
   - add kunit helpers to create a device

  sched:
   - module param for scheduling policy
   - refcounting fix

  buddy:
   - add back random seed log

  ttm:
   - convert ttm_resource to size_t
   - optimize pool allocations

  edid:
   - HFVSDB parsing support fixes
   - logging/debug improvements
   - DSC quirks

  dma-buf:
   - Add unlocked vmap and attachment mapping
   - move drivers to common locking convention
   - locking improvements

  firmware:
   - new API for rPI firmware and vc4

  xilinx:
   - zynqmp: displayport bridge support
   - dpsub fix

  bridge:
   - adv7533: Remove dynamic lane switching
   - it6505: Runtime PM support, sync improvements
   - ps8640: Handle AUX defer messages
   - tc358775: Drop soft-reset over I2C

  panel:
   - panel-edp: Add INX N116BGE-EA2 C2 and C4 support.
   - Jadard JD9365DA-H3
   - NewVision NV3051D

  amdgpu:
   - DCN support on ARM
   - DCN 2.1 secure display
   - Sienna Cichlid mode2 reset fixes
   - new GC 11.x firmware versions
   - drop AMD specific DSC workarounds in favour of drm code
   - clang warning fixes
   - scheduler rework
   - SR-IOV fixes
   - GPUVM locking fixes
   - fix memory leak in CS IOCTL error path
   - flexible array updates
   - enable new GC/PSP/SMU/NBIO IP
   - GFX preemption support for gfx9

  amdkfd:
   - cache size fixes
   - userptr fixes
   - enable cooperative launch on gfx 10.3
   - enable GC 11.0.4 KFD support

  radeon:
   - replace kmap with kmap_local_page
   - ACPI ref count fix
   - HDA audio notifier support

  i915:
   - DG2 enabled by default
   - MTL enablement work
   - hotplug refactoring
   - VBT improvements
   - Display and watermark refactoring
   - ADL-P workaround
   - temp disable runtime_pm for discrete-
   - fix for A380 as a secondary GPU
   - Wa_18017747507 for DG2
   - CS timestamp support fixes for gen5 and earlier
   - never purge busy TTM objects
   - use i915_sg_dma_sizes for all backends
   - demote GuC kernel contexts to normal priority
   - gvt: refactor for new MDEV interface
   - enable DC power states on eDP ports
   - fix gen 2/3 workarounds

  nouveau:
   - fix page fault handling
   - Ampere acceleration support
   - driver stability improvements
   - nva3 backlight support

  msm:
   - MSM_INFO_GET_FLAGS support
   - DPU: XR30 and P010 image formats
   - Qualcomm SM6115 support
   - DSI PHY support for QCM2290
   - HDMI: refactored dev init path
   - remove exclusive-fence hack
   - fix speed-bin detection
   - enable clamp to idle on 7c3
   - improved hangcheck detection

  vmwgfx:
   - fb and cursor refactoring
   - convert to generic hashtable
   - cursor improvements

  etnaviv:
   - hw workarounds
   - softpin MMU fixes

  ast:
   - atomic gamma LUT support
   - convert to SHMEM

  lcdif:
   - support YUV planes
   - Increase DMA burst size
   - FIFO threshold tuning

  meson:
   - fix return type of cvbs mode_valid

  mgag200:
   - fix PLL setup on some revisions

  sun4i:
   - A100 and D1 support

  udl:
   - modesetting improvements
   - hot unplug support

  vc4:
   - support PAL-M
   - fix regression preventing 4K @ 60Hz
   - fix NULL ptr deref

  v3d:
   - switch to drm managed resources

  renesas:
   - RZ/G2L DSI support
   - DU Kconfig cleanup

  mediatek:
   - fixup dpi and hdmi
   - MT8188 dpi support
   - MT8195 AFBC support

  tegra:
   - NVDEC hardware on Tegra234 SoC

  hdlcd:
   - switch to drm managed resources

  ingenic:
   - fix registration error path

  hisilicon:
   - convert to drm_mode_init

  maildp:
   - use managed resources

  mtk:
   - use drm_mode_init

  rockchip:
   - use drm_mode_copy"

* tag 'drm-next-2022-12-13' of git://anongit.freedesktop.org/drm/drm: (1397 commits)
  drm/amdgpu: fix mmhub register base coding error
  drm/amdgpu: add tmz support for GC IP v11.0.4
  drm/amdgpu: enable GFX Clock Gating control for GC IP v11.0.4
  drm/amdgpu: enable GFX Power Gating for GC IP v11.0.4
  drm/amdgpu: enable GFX IP v11.0.4 CG support
  drm/amdgpu: Make amdgpu_ring_mux functions as static
  drm/amdgpu: generally allow over-commit during BO allocation
  drm/amd/display: fix array index out of bound error in DCN32 DML
  drm/amd/display: 3.2.215
  drm/amd/display: set optimized required for comp buf changes
  drm/amd/display: Add debug option to skip PSR CRTC disable
  drm/amd/display: correct DML calc error of UrgentLatency
  drm/amd/display: correct static_screen_event_mask
  drm/amd/display: Ensure commit_streams returns the DC return code
  drm/amd/display: read invalid ddc pin status cause engine busy
  drm/amd/display: Bypass DET swath fill check for max clocks
  drm/amd/display: Disable uclk pstate for subvp pipes
  drm/amd/display: Fix DCN2.1 default DSC clocks
  drm/amd/display: Enable dp_hdmi21_pcon support
  drm/amd/display: prevent seamless boot on displays that don't have the preferred dig
  ...

11 files changed, 465 insertions, 737 deletions

diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
index 8bfdfd062ff6..3251f4783ba1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.c
@@ -50,16 +50,6 @@ static inline unsigned int get_and_inc_gpu_processor_id(
 	return current_id;
 }
 
-/* Static table to describe GPU Cache information */
-struct kfd_gpu_cache_info {
-	uint32_t	cache_size;
-	uint32_t	cache_level;
-	uint32_t	flags;
-	/* Indicates how many Compute Units share this cache
-	 * within a SA. Value = 1 indicates the cache is not shared
-	 */
-	uint32_t	num_cu_shared;
-};
 
 static struct kfd_gpu_cache_info kaveri_cache_info[] = {
 	{
@@ -891,6 +881,54 @@ static struct kfd_gpu_cache_info gc_10_3_6_cache_info[] = {
 	},
 };
 
+static struct kfd_gpu_cache_info dummy_cache_info[] = {
+	{
+		/* TCP L1 Cache per CU */
+		.cache_size = 16,
+		.cache_level = 1,
+		.flags = (CRAT_CACHE_FLAGS_ENABLED |
+				CRAT_CACHE_FLAGS_DATA_CACHE |
+				CRAT_CACHE_FLAGS_SIMD_CACHE),
+		.num_cu_shared = 1,
+	},
+	{
+		/* Scalar L1 Instruction Cache per SQC */
+		.cache_size = 32,
+		.cache_level = 1,
+		.flags = (CRAT_CACHE_FLAGS_ENABLED |
+				CRAT_CACHE_FLAGS_INST_CACHE |
+				CRAT_CACHE_FLAGS_SIMD_CACHE),
+		.num_cu_shared = 2,
+	},
+	{
+		/* Scalar L1 Data Cache per SQC */
+		.cache_size = 16,
+		.cache_level = 1,
+		.flags = (CRAT_CACHE_FLAGS_ENABLED |
+				CRAT_CACHE_FLAGS_DATA_CACHE |
+				CRAT_CACHE_FLAGS_SIMD_CACHE),
+		.num_cu_shared = 2,
+	},
+	{
+		/* GL1 Data Cache per SA */
+		.cache_size = 128,
+		.cache_level = 1,
+		.flags = (CRAT_CACHE_FLAGS_ENABLED |
+				CRAT_CACHE_FLAGS_DATA_CACHE |
+				CRAT_CACHE_FLAGS_SIMD_CACHE),
+		.num_cu_shared = 6,
+	},
+	{
+		/* L2 Data Cache per GPU (Total Tex Cache) */
+		.cache_size = 2048,
+		.cache_level = 2,
+		.flags = (CRAT_CACHE_FLAGS_ENABLED |
+				CRAT_CACHE_FLAGS_DATA_CACHE |
+				CRAT_CACHE_FLAGS_SIMD_CACHE),
+		.num_cu_shared = 6,
+	},
+};
+
 static void kfd_populated_cu_info_cpu(struct kfd_topology_device *dev,
 		struct crat_subtype_computeunit *cu)
 {
@@ -1071,8 +1109,12 @@ static int kfd_parse_subtype_cache(struct crat_subtype_cache *cache,
 			props->cachelines_per_tag = cache->lines_per_tag;
 			props->cache_assoc = cache->associativity;
 			props->cache_latency = cache->cache_latency;
+
 			memcpy(props->sibling_map, cache->sibling_map,
-					sizeof(props->sibling_map));
+					CRAT_SIBLINGMAP_SIZE);
+
+			/* set the sibling_map_size as 32 for CRAT from ACPI */
+			props->sibling_map_size = CRAT_SIBLINGMAP_SIZE;
 
 			if (cache->flags & CRAT_CACHE_FLAGS_DATA_CACHE)
 				props->cache_type |= HSA_CACHE_TYPE_DATA;
@@ -1083,7 +1125,6 @@ static int kfd_parse_subtype_cache(struct crat_subtype_cache *cache,
 			if (cache->flags & CRAT_CACHE_FLAGS_SIMD_CACHE)
 				props->cache_type |= HSA_CACHE_TYPE_HSACU;
 
-			dev->cache_count++;
 			dev->node_props.caches_count++;
 			list_add_tail(&props->list, &dev->cache_props);
 
@@ -1291,125 +1332,6 @@ err:
 	return ret;
 }
 
-/* Helper function. See kfd_fill_gpu_cache_info for parameter description */
-static int fill_in_l1_pcache(struct crat_subtype_cache *pcache,
-				struct kfd_gpu_cache_info *pcache_info,
-				struct kfd_cu_info *cu_info,
-				int mem_available,
-				int cu_bitmask,
-				int cache_type, unsigned int cu_processor_id,
-				int cu_block)
-{
-	unsigned int cu_sibling_map_mask;
-	int first_active_cu;
-
-	/* First check if enough memory is available */
-	if (sizeof(struct crat_subtype_cache) > mem_available)
-		return -ENOMEM;
-
-	cu_sibling_map_mask = cu_bitmask;
-	cu_sibling_map_mask >>= cu_block;
-	cu_sibling_map_mask &=
-		((1 << pcache_info[cache_type].num_cu_shared) - 1);
-	first_active_cu = ffs(cu_sibling_map_mask);
-
-	/* CU could be inactive. In case of shared cache find the first active
-	 * CU. and incase of non-shared cache check if the CU is inactive. If
-	 * inactive active skip it
-	 */
-	if (first_active_cu) {
-		memset(pcache, 0, sizeof(struct crat_subtype_cache));
-		pcache->type = CRAT_SUBTYPE_CACHE_AFFINITY;
-		pcache->length = sizeof(struct crat_subtype_cache);
-		pcache->flags = pcache_info[cache_type].flags;
-		pcache->processor_id_low = cu_processor_id
-					 + (first_active_cu - 1);
-		pcache->cache_level = pcache_info[cache_type].cache_level;
-		pcache->cache_size = pcache_info[cache_type].cache_size;
-
-		/* Sibling map is w.r.t processor_id_low, so shift out
-		 * inactive CU
-		 */
-		cu_sibling_map_mask =
-			cu_sibling_map_mask >> (first_active_cu - 1);
-
-		pcache->sibling_map[0] = (uint8_t)(cu_sibling_map_mask & 0xFF);
-		pcache->sibling_map[1] =
-				(uint8_t)((cu_sibling_map_mask >> 8) & 0xFF);
-		pcache->sibling_map[2] =
-				(uint8_t)((cu_sibling_map_mask >> 16) & 0xFF);
-		pcache->sibling_map[3] =
-				(uint8_t)((cu_sibling_map_mask >> 24) & 0xFF);
-		return 0;
-	}
-	return 1;
-}
-
-/* Helper function. See kfd_fill_gpu_cache_info for parameter description */
-static int fill_in_l2_l3_pcache(struct crat_subtype_cache *pcache,
-				struct kfd_gpu_cache_info *pcache_info,
-				struct kfd_cu_info *cu_info,
-				int mem_available,
-				int cache_type, unsigned int cu_processor_id)
-{
-	unsigned int cu_sibling_map_mask;
-	int first_active_cu;
-	int i, j, k;
-
-	/* First check if enough memory is available */
-	if (sizeof(struct crat_subtype_cache) > mem_available)
-		return -ENOMEM;
-
-	cu_sibling_map_mask = cu_info->cu_bitmap[0][0];
-	cu_sibling_map_mask &=
-		((1 << pcache_info[cache_type].num_cu_shared) - 1);
-	first_active_cu = ffs(cu_sibling_map_mask);
-
-	/* CU could be inactive. In case of shared cache find the first active
-	 * CU. and incase of non-shared cache check if the CU is inactive. If
-	 * inactive active skip it
-	 */
-	if (first_active_cu) {
-		memset(pcache, 0, sizeof(struct crat_subtype_cache));
-		pcache->type = CRAT_SUBTYPE_CACHE_AFFINITY;
-		pcache->length = sizeof(struct crat_subtype_cache);
-		pcache->flags = pcache_info[cache_type].flags;
-		pcache->processor_id_low = cu_processor_id
-					 + (first_active_cu - 1);
-		pcache->cache_level = pcache_info[cache_type].cache_level;
-		pcache->cache_size = pcache_info[cache_type].cache_size;
-
-		/* Sibling map is w.r.t processor_id_low, so shift out
-		 * inactive CU
-		 */
-		cu_sibling_map_mask =
-			cu_sibling_map_mask >> (first_active_cu - 1);
-		k = 0;
-		for (i = 0; i < cu_info->num_shader_engines; i++) {
-			for (j = 0; j < cu_info->num_shader_arrays_per_engine;
-				j++) {
-				pcache->sibling_map[k] =
-				 (uint8_t)(cu_sibling_map_mask & 0xFF);
-				pcache->sibling_map[k+1] =
-				 (uint8_t)((cu_sibling_map_mask >> 8) & 0xFF);
-				pcache->sibling_map[k+2] =
-				 (uint8_t)((cu_sibling_map_mask >> 16) & 0xFF);
-				pcache->sibling_map[k+3] =
-				 (uint8_t)((cu_sibling_map_mask >> 24) & 0xFF);
-				k += 4;
-				cu_sibling_map_mask =
-					cu_info->cu_bitmap[i % 4][j + i / 4];
-				cu_sibling_map_mask &= (
-				 (1 << pcache_info[cache_type].num_cu_shared)
-				 - 1);
-			}
-		}
-		return 0;
-	}
-	return 1;
-}
-
-#define KFD_MAX_CACHE_TYPES 6
 
 static int kfd_fill_gpu_cache_info_from_gfx_config(struct kfd_dev *kdev,
 						   struct kfd_gpu_cache_info *pcache_info)
@@ -1483,228 +1405,134 @@ static int kfd_fill_gpu_cache_info_from_gfx_config(struct kfd_dev *kdev,
 	return i;
 }
 
-/* kfd_fill_gpu_cache_info - Fill GPU cache info using kfd_gpu_cache_info
- * tables
- *
- *	@kdev - [IN] GPU device
- *	@gpu_processor_id - [IN] GPU processor ID to which these caches
- *			    associate
- *	@available_size - [IN] Amount of memory available in pcache
- *	@cu_info - [IN] Compute Unit info obtained from KGD
- *	@pcache - [OUT] memory into which cache data is to be filled in.
- *	@size_filled - [OUT] amount of data used up in pcache.
- *	@num_of_entries - [OUT] number of caches added
- */
-static int kfd_fill_gpu_cache_info(struct kfd_dev *kdev,
-			int gpu_processor_id,
-			int available_size,
-			struct kfd_cu_info *cu_info,
-			struct crat_subtype_cache *pcache,
-			int *size_filled,
-			int *num_of_entries)
+int kfd_get_gpu_cache_info(struct kfd_dev *kdev, struct kfd_gpu_cache_info **pcache_info)
 {
-	struct kfd_gpu_cache_info *pcache_info;
-	struct kfd_gpu_cache_info cache_info[KFD_MAX_CACHE_TYPES];
 	int num_of_cache_types = 0;
-	int i, j, k;
-	int ct = 0;
-	int mem_available = available_size;
-	unsigned int cu_processor_id;
-	int ret;
-	unsigned int num_cu_shared;
 
 	switch (kdev->adev->asic_type) {
 	case CHIP_KAVERI:
-		pcache_info = kaveri_cache_info;
+		*pcache_info = kaveri_cache_info;
 		num_of_cache_types = ARRAY_SIZE(kaveri_cache_info);
 		break;
 	case CHIP_HAWAII:
-		pcache_info = hawaii_cache_info;
+		*pcache_info = hawaii_cache_info;
 		num_of_cache_types = ARRAY_SIZE(hawaii_cache_info);
 		break;
 	case CHIP_CARRIZO:
-		pcache_info = carrizo_cache_info;
+		*pcache_info = carrizo_cache_info;
 		num_of_cache_types = ARRAY_SIZE(carrizo_cache_info);
 		break;
 	case CHIP_TONGA:
-		pcache_info = tonga_cache_info;
+		*pcache_info = tonga_cache_info;
 		num_of_cache_types = ARRAY_SIZE(tonga_cache_info);
 		break;
 	case CHIP_FIJI:
-		pcache_info = fiji_cache_info;
+		*pcache_info = fiji_cache_info;
 		num_of_cache_types = ARRAY_SIZE(fiji_cache_info);
 		break;
 	case CHIP_POLARIS10:
-		pcache_info = polaris10_cache_info;
+		*pcache_info = polaris10_cache_info;
 		num_of_cache_types = ARRAY_SIZE(polaris10_cache_info);
 		break;
 	case CHIP_POLARIS11:
-		pcache_info = polaris11_cache_info;
+		*pcache_info = polaris11_cache_info;
 		num_of_cache_types = ARRAY_SIZE(polaris11_cache_info);
 		break;
 	case CHIP_POLARIS12:
-		pcache_info = polaris12_cache_info;
+		*pcache_info = polaris12_cache_info;
 		num_of_cache_types = ARRAY_SIZE(polaris12_cache_info);
 		break;
 	case CHIP_VEGAM:
-		pcache_info = vegam_cache_info;
+		*pcache_info = vegam_cache_info;
 		num_of_cache_types = ARRAY_SIZE(vegam_cache_info);
 		break;
 	default:
 		switch (KFD_GC_VERSION(kdev)) {
 		case IP_VERSION(9, 0, 1):
-			pcache_info = vega10_cache_info;
+			*pcache_info = vega10_cache_info;
 			num_of_cache_types = ARRAY_SIZE(vega10_cache_info);
 			break;
 		case IP_VERSION(9, 2, 1):
-			pcache_info = vega12_cache_info;
+			*pcache_info = vega12_cache_info;
 			num_of_cache_types = ARRAY_SIZE(vega12_cache_info);
 			break;
 		case IP_VERSION(9, 4, 0):
 		case IP_VERSION(9, 4, 1):
-			pcache_info = vega20_cache_info;
+			*pcache_info = vega20_cache_info;
 			num_of_cache_types = ARRAY_SIZE(vega20_cache_info);
 			break;
 		case IP_VERSION(9, 4, 2):
-			pcache_info = aldebaran_cache_info;
+			*pcache_info = aldebaran_cache_info;
 			num_of_cache_types = ARRAY_SIZE(aldebaran_cache_info);
 			break;
 		case IP_VERSION(9, 1, 0):
 		case IP_VERSION(9, 2, 2):
-			pcache_info = raven_cache_info;
+			*pcache_info = raven_cache_info;
 			num_of_cache_types = ARRAY_SIZE(raven_cache_info);
 			break;
 		case IP_VERSION(9, 3, 0):
-			pcache_info = renoir_cache_info;
+			*pcache_info = renoir_cache_info;
 			num_of_cache_types = ARRAY_SIZE(renoir_cache_info);
 			break;
 		case IP_VERSION(10, 1, 10):
 		case IP_VERSION(10, 1, 2):
 		case IP_VERSION(10, 1, 3):
 		case IP_VERSION(10, 1, 4):
-			pcache_info = navi10_cache_info;
+			*pcache_info = navi10_cache_info;
 			num_of_cache_types = ARRAY_SIZE(navi10_cache_info);
 			break;
 		case IP_VERSION(10, 1, 1):
-			pcache_info = navi14_cache_info;
+			*pcache_info = navi14_cache_info;
 			num_of_cache_types = ARRAY_SIZE(navi14_cache_info);
 			break;
 		case IP_VERSION(10, 3, 0):
-			pcache_info = sienna_cichlid_cache_info;
+			*pcache_info = sienna_cichlid_cache_info;
 			num_of_cache_types = ARRAY_SIZE(sienna_cichlid_cache_info);
 			break;
 		case IP_VERSION(10, 3, 2):
-			pcache_info = navy_flounder_cache_info;
+			*pcache_info = navy_flounder_cache_info;
 			num_of_cache_types = ARRAY_SIZE(navy_flounder_cache_info);
 			break;
 		case IP_VERSION(10, 3, 4):
-			pcache_info = dimgrey_cavefish_cache_info;
+			*pcache_info = dimgrey_cavefish_cache_info;
 			num_of_cache_types = ARRAY_SIZE(dimgrey_cavefish_cache_info);
 			break;
 		case IP_VERSION(10, 3, 1):
-			pcache_info = vangogh_cache_info;
+			*pcache_info = vangogh_cache_info;
 			num_of_cache_types = ARRAY_SIZE(vangogh_cache_info);
 			break;
 		case IP_VERSION(10, 3, 5):
-			pcache_info = beige_goby_cache_info;
+			*pcache_info = beige_goby_cache_info;
 			num_of_cache_types = ARRAY_SIZE(beige_goby_cache_info);
 			break;
 		case IP_VERSION(10, 3, 3):
-			pcache_info = yellow_carp_cache_info;
+			*pcache_info = yellow_carp_cache_info;
 			num_of_cache_types = ARRAY_SIZE(yellow_carp_cache_info);
 			break;
 		case IP_VERSION(10, 3, 6):
-			pcache_info = gc_10_3_6_cache_info;
+			*pcache_info = gc_10_3_6_cache_info;
 			num_of_cache_types = ARRAY_SIZE(gc_10_3_6_cache_info);
 			break;
 		case IP_VERSION(10, 3, 7):
-			pcache_info = gfx1037_cache_info;
+			*pcache_info = gfx1037_cache_info;
 			num_of_cache_types = ARRAY_SIZE(gfx1037_cache_info);
 			break;
 		case IP_VERSION(11, 0, 0):
 		case IP_VERSION(11, 0, 1):
 		case IP_VERSION(11, 0, 2):
 		case IP_VERSION(11, 0, 3):
-			pcache_info = cache_info;
+		case IP_VERSION(11, 0, 4):
 			num_of_cache_types =
-				kfd_fill_gpu_cache_info_from_gfx_config(kdev, pcache_info);
+				kfd_fill_gpu_cache_info_from_gfx_config(kdev, *pcache_info);
 			break;
 		default:
-			return -EINVAL;
-		}
-	}
-
-	*size_filled = 0;
-	*num_of_entries = 0;
-
-	/* For each type of cache listed in the kfd_gpu_cache_info table,
-	 * go through all available Compute Units.
-	 * The [i,j,k] loop will
-	 *		if kfd_gpu_cache_info.num_cu_shared = 1
-	 *			will parse through all available CU
-	 *		If (kfd_gpu_cache_info.num_cu_shared != 1)
-	 *			then it will consider only one CU from
-	 *			the shared unit
-	 */
-
-	for (ct = 0; ct < num_of_cache_types; ct++) {
-	  cu_processor_id = gpu_processor_id;
-	  if (pcache_info[ct].cache_level == 1) {
-	    for (i = 0; i < cu_info->num_shader_engines; i++) {
-	      for (j = 0; j < cu_info->num_shader_arrays_per_engine; j++) {
-	        for (k = 0; k < cu_info->num_cu_per_sh;
-		  k += pcache_info[ct].num_cu_shared) {
-		  ret = fill_in_l1_pcache(pcache,
-					pcache_info,
-					cu_info,
-					mem_available,
-					cu_info->cu_bitmap[i % 4][j + i / 4],
-					ct,
-					cu_processor_id,
-					k);
-
-		  if (ret < 0)
+			*pcache_info = dummy_cache_info;
+			num_of_cache_types = ARRAY_SIZE(dummy_cache_info);
+			pr_warn("dummy cache info is used temporarily and real cache info need update later.\n");
 			break;
-
-		  if (!ret) {
-				pcache++;
-				(*num_of_entries)++;
-				mem_available -= sizeof(*pcache);
-				(*size_filled) += sizeof(*pcache);
-		  }
-
-		  /* Move to next CU block */
-		  num_cu_shared = ((k + pcache_info[ct].num_cu_shared) <=
-					cu_info->num_cu_per_sh) ?
-					pcache_info[ct].num_cu_shared :
-					(cu_info->num_cu_per_sh - k);
-		  cu_processor_id += num_cu_shared;
 		}
-	      }
-	    }
-	  } else {
-			ret = fill_in_l2_l3_pcache(pcache,
-				pcache_info,
-				cu_info,
-				mem_available,
-				ct,
-				cu_processor_id);
-
-			if (ret < 0)
-				break;
-
-			if (!ret) {
-				pcache++;
-				(*num_of_entries)++;
-				mem_available -= sizeof(*pcache);
-				(*size_filled) += sizeof(*pcache);
-			}
-	  }
 	}
-
-	pr_debug("Added [%d] GPU cache entries\n", *num_of_entries);
-
-	return 0;
+	return num_of_cache_types;
 }
 
 static bool kfd_ignore_crat(void)
@@ -2063,8 +1891,8 @@ static void kfd_find_numa_node_in_srat(struct kfd_dev *kdev)
 	struct acpi_table_header *table_header = NULL;
 	struct acpi_subtable_header *sub_header = NULL;
 	unsigned long table_end, subtable_len;
-	u32 pci_id = pci_domain_nr(kdev->pdev->bus) << 16 |
-			pci_dev_id(kdev->pdev);
+	u32 pci_id = pci_domain_nr(kdev->adev->pdev->bus) << 16 |
+			pci_dev_id(kdev->adev->pdev);
 	u32 bdf;
 	acpi_status status;
 	struct acpi_srat_cpu_affinity *cpu;
@@ -2139,7 +1967,7 @@ static void kfd_find_numa_node_in_srat(struct kfd_dev *kdev)
 		numa_node = 0;
 
 	if (numa_node != NUMA_NO_NODE)
-		set_dev_node(&kdev->pdev->dev, numa_node);
+		set_dev_node(&kdev->adev->pdev->dev, numa_node);
 }
 #endif
 
@@ -2200,14 +2028,14 @@ static int kfd_fill_gpu_direct_io_link_to_cpu(int *avail_size,
 	sub_type_hdr->proximity_domain_from = proximity_domain;
 
 #ifdef CONFIG_ACPI_NUMA
-	if (kdev->pdev->dev.numa_node == NUMA_NO_NODE)
+	if (kdev->adev->pdev->dev.numa_node == NUMA_NO_NODE)
 		kfd_find_numa_node_in_srat(kdev);
 #endif
 #ifdef CONFIG_NUMA
-	if (kdev->pdev->dev.numa_node == NUMA_NO_NODE)
+	if (kdev->adev->pdev->dev.numa_node == NUMA_NO_NODE)
 		sub_type_hdr->proximity_domain_to = 0;
 	else
-		sub_type_hdr->proximity_domain_to = kdev->pdev->dev.numa_node;
+		sub_type_hdr->proximity_domain_to = kdev->adev->pdev->dev.numa_node;
 #else
 	sub_type_hdr->proximity_domain_to = 0;
 #endif
@@ -2263,8 +2091,6 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image,
 	struct kfd_cu_info cu_info;
 	int avail_size = *size;
 	uint32_t total_num_of_cu;
-	int num_of_cache_entries = 0;
-	int cache_mem_filled = 0;
 	uint32_t nid = 0;
 	int ret = 0;
 
@@ -2365,31 +2191,12 @@ static int kfd_create_vcrat_image_gpu(void *pcrat_image,
 	crat_table->length += sizeof(struct crat_subtype_memory);
 	crat_table->total_entries++;
 
-	/* TODO: Fill in cache information. This information is NOT readily
-	 * available in KGD
-	 */
-	sub_type_hdr = (typeof(sub_type_hdr))((char *)sub_type_hdr +
-		sub_type_hdr->length);
-	ret = kfd_fill_gpu_cache_info(kdev, cu->processor_id_low,
-				avail_size,
-				&cu_info,
-				(struct crat_subtype_cache *)sub_type_hdr,
-				&cache_mem_filled,
-				&num_of_cache_entries);
-
-	if (ret < 0)
-		return ret;
-
-	crat_table->length += cache_mem_filled;
-	crat_table->total_entries += num_of_cache_entries;
-	avail_size -= cache_mem_filled;
-
 	/* Fill in Subtype: IO_LINKS
 	 *  Only direct links are added here which is Link from GPU to
 	 *  its NUMA node. Indirect links are added by userspace.
 	 */
 	sub_type_hdr = (typeof(sub_type_hdr))((char *)sub_type_hdr +
-		cache_mem_filled);
+		sub_type_hdr->length);
 	ret = kfd_fill_gpu_direct_io_link_to_cpu(&avail_size, kdev,
 		(struct crat_subtype_iolink *)sub_type_hdr, proximity_domain);
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_crat.h b/drivers/gpu/drm/amd/amdkfd/kfd_crat.h
index 482ba84a728d..8d1e8ba58dee 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_crat.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_crat.h
@@ -29,11 +29,10 @@
 #pragma pack(1)
 
 /*
- * 4CC signature values for the CRAT and CDIT ACPI tables
+ * 4CC signature value for the CRAT ACPI table
  */
 
 #define CRAT_SIGNATURE	"CRAT"
-#define CDIT_SIGNATURE	"CDIT"
 
 /*
  * Component Resource Association Table (CRAT)
@@ -292,31 +291,22 @@ struct crat_subtype_generic {
 	uint32_t	flags;
 };
 
-/*
- * Component Locality Distance Information Table (CDIT)
- */
-#define CDIT_OEMID_LENGTH	6
-#define CDIT_OEMTABLEID_LENGTH	8
-
-struct cdit_header {
-	uint32_t	signature;
-	uint32_t	length;
-	uint8_t		revision;
-	uint8_t		checksum;
-	uint8_t		oem_id[CDIT_OEMID_LENGTH];
-	uint8_t		oem_table_id[CDIT_OEMTABLEID_LENGTH];
-	uint32_t	oem_revision;
-	uint32_t	creator_id;
-	uint32_t	creator_revision;
-	uint32_t	total_entries;
-	uint16_t	num_domains;
-	uint8_t		entry[1];
-};
-
 #pragma pack()
 
 struct kfd_dev;
 
+/* Static table to describe GPU Cache information */
+struct kfd_gpu_cache_info {
+	uint32_t	cache_size;
+	uint32_t	cache_level;
+	uint32_t	flags;
+	/* Indicates how many Compute Units share this cache
+	 * within a SA. Value = 1 indicates the cache is not shared
+	 */
+	uint32_t	num_cu_shared;
+};
+int kfd_get_gpu_cache_info(struct kfd_dev *kdev, struct kfd_gpu_cache_info **pcache_info);
+
 int kfd_create_crat_image_acpi(void **crat_image, size_t *size);
 void kfd_destroy_crat_image(void *crat_image);
 int kfd_parse_crat_table(void *crat_image, struct list_head *device_list,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_device.c b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
index 65a1d4f9004b..b8936340742b 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_device.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_device.c
@@ -153,6 +153,7 @@ static void kfd_device_info_set_event_interrupt_class(struct kfd_dev *kfd)
 	case IP_VERSION(11, 0, 1):
 	case IP_VERSION(11, 0, 2):
 	case IP_VERSION(11, 0, 3):
+	case IP_VERSION(11, 0, 4):
 		kfd->device_info.event_interrupt_class = &event_interrupt_class_v11;
 		break;
 	default:
@@ -227,7 +228,6 @@ struct kfd_dev *kgd2kfd_probe(struct amdgpu_device *adev, bool vf)
 {
 	struct kfd_dev *kfd = NULL;
 	const struct kfd2kgd_calls *f2g = NULL;
-	struct pci_dev *pdev = adev->pdev;
 	uint32_t gfx_target_version = 0;
 
 	switch (adev->asic_type) {
@@ -395,6 +395,7 @@ struct kfd_dev *kgd2kfd_probe(struct amdgpu_device *adev, bool vf)
 			f2g = &gfx_v11_kfd2kgd;
 			break;
 		case IP_VERSION(11, 0, 1):
+		case IP_VERSION(11, 0, 4):
 			gfx_target_version = 110003;
 			f2g = &gfx_v11_kfd2kgd;
 			break;
@@ -429,7 +430,6 @@ struct kfd_dev *kgd2kfd_probe(struct amdgpu_device *adev, bool vf)
 
 	kfd->adev = adev;
 	kfd_device_info_init(kfd, vf, gfx_target_version);
-	kfd->pdev = pdev;
 	kfd->init_complete = false;
 	kfd->kfd2kgd = f2g;
 	atomic_set(&kfd->compute_profile, 0);
@@ -497,7 +497,10 @@ static int kfd_gws_init(struct kfd_dev *kfd)
 		(KFD_GC_VERSION(kfd) == IP_VERSION(9, 4, 1)
 			&& kfd->mec2_fw_version >= 0x30)   ||
 		(KFD_GC_VERSION(kfd) == IP_VERSION(9, 4, 2)
-			&& kfd->mec2_fw_version >= 0x28))))
+			&& kfd->mec2_fw_version >= 0x28) ||
+		(KFD_GC_VERSION(kfd) >= IP_VERSION(10, 3, 0)
+			&& KFD_GC_VERSION(kfd) < IP_VERSION(11, 0, 0)
+			&& kfd->mec2_fw_version >= 0x6b))))
 		ret = amdgpu_amdkfd_alloc_gws(kfd->adev,
 				kfd->adev->gds.gws_size, &kfd->gws);
 
@@ -511,12 +514,10 @@ static void kfd_smi_init(struct kfd_dev *dev)
 }
 
 bool kgd2kfd_device_init(struct kfd_dev *kfd,
-			 struct drm_device *ddev,
 			 const struct kgd2kfd_shared_resources *gpu_resources)
 {
 	unsigned int size, map_process_packet_size;
 
-	kfd->ddev = ddev;
 	kfd->mec_fw_version = amdgpu_amdkfd_get_fw_version(kfd->adev,
 			KGD_ENGINE_MEC1);
 	kfd->mec2_fw_version = amdgpu_amdkfd_get_fw_version(kfd->adev,
@@ -541,7 +542,7 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 	     kfd->mec_fw_version < kfd->device_info.no_atomic_fw_version)) {
 		dev_info(kfd_device,
 			 "skipped device %x:%x, PCI rejects atomics %d<%d\n",
-			 kfd->pdev->vendor, kfd->pdev->device,
+			 kfd->adev->pdev->vendor, kfd->adev->pdev->device,
 			 kfd->mec_fw_version,
 			 kfd->device_info.no_atomic_fw_version);
 		return false;
@@ -650,8 +651,8 @@ bool kgd2kfd_device_init(struct kfd_dev *kfd,
 	kfd_smi_init(kfd);
 
 	kfd->init_complete = true;
-	dev_info(kfd_device, "added device %x:%x\n", kfd->pdev->vendor,
-		 kfd->pdev->device);
+	dev_info(kfd_device, "added device %x:%x\n", kfd->adev->pdev->vendor,
+		 kfd->adev->pdev->device);
 
 	pr_debug("Starting kfd with the following scheduling policy %d\n",
 		kfd->dqm->sched_policy);
@@ -676,7 +677,7 @@ alloc_gtt_mem_failure:
 		amdgpu_amdkfd_free_gws(kfd->adev, kfd->gws);
 	dev_err(kfd_device,
 		"device %x:%x NOT added due to errors\n",
-		kfd->pdev->vendor, kfd->pdev->device);
+		kfd->adev->pdev->vendor, kfd->adev->pdev->device);
 out:
 	return kfd->init_complete;
 }
@@ -789,7 +790,7 @@ int kgd2kfd_resume_iommu(struct kfd_dev *kfd)
 	if (err)
 		dev_err(kfd_device,
 			"Failed to resume IOMMU for device %x:%x\n",
-			kfd->pdev->vendor, kfd->pdev->device);
+			kfd->adev->pdev->vendor, kfd->adev->pdev->device);
 	return err;
 }
 
@@ -801,7 +802,7 @@ static int kfd_resume(struct kfd_dev *kfd)
 	if (err)
 		dev_err(kfd_device,
 			"Error starting queue manager for device %x:%x\n",
-			kfd->pdev->vendor, kfd->pdev->device);
+			kfd->adev->pdev->vendor, kfd->adev->pdev->device);
 
 	return err;
 }
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
index fbd0afe4da42..ec1bf611624e 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_iommu.c
@@ -49,7 +49,7 @@ int kfd_iommu_check_device(struct kfd_dev *kfd)
 		return -ENODEV;
 
 	iommu_info.flags = 0;
-	err = amd_iommu_device_info(kfd->pdev, &iommu_info);
+	err = amd_iommu_device_info(kfd->adev->pdev, &iommu_info);
 	if (err)
 		return err;
 
@@ -71,7 +71,7 @@ int kfd_iommu_device_init(struct kfd_dev *kfd)
 		return 0;
 
 	iommu_info.flags = 0;
-	err = amd_iommu_device_info(kfd->pdev, &iommu_info);
+	err = amd_iommu_device_info(kfd->adev->pdev, &iommu_info);
 	if (err < 0) {
 		dev_err(kfd_device,
 			"error getting iommu info. is the iommu enabled?\n");
@@ -121,7 +121,7 @@ int kfd_iommu_bind_process_to_device(struct kfd_process_device *pdd)
 		return -EINVAL;
 	}
 
-	err = amd_iommu_bind_pasid(dev->pdev, p->pasid, p->lead_thread);
+	err = amd_iommu_bind_pasid(dev->adev->pdev, p->pasid, p->lead_thread);
 	if (!err)
 		pdd->bound = PDD_BOUND;
 
@@ -139,7 +139,8 @@ void kfd_iommu_unbind_process(struct kfd_process *p)
 
 	for (i = 0; i < p->n_pdds; i++)
 		if (p->pdds[i]->bound == PDD_BOUND)
-			amd_iommu_unbind_pasid(p->pdds[i]->dev->pdev, p->pasid);
+			amd_iommu_unbind_pasid(p->pdds[i]->dev->adev->pdev,
+					       p->pasid);
 }
 
 /* Callback for process shutdown invoked by the IOMMU driver */
@@ -222,7 +223,7 @@ static int kfd_bind_processes_to_device(struct kfd_dev *kfd)
 			continue;
 		}
 
-		err = amd_iommu_bind_pasid(kfd->pdev, p->pasid,
+		err = amd_iommu_bind_pasid(kfd->adev->pdev, p->pasid,
 				p->lead_thread);
 		if (err < 0) {
 			pr_err("Unexpected pasid 0x%x binding failure\n",
@@ -282,9 +283,9 @@ void kfd_iommu_suspend(struct kfd_dev *kfd)
 
 	kfd_unbind_processes_from_device(kfd);
 
-	amd_iommu_set_invalidate_ctx_cb(kfd->pdev, NULL);
-	amd_iommu_set_invalid_ppr_cb(kfd->pdev, NULL);
-	amd_iommu_free_device(kfd->pdev);
+	amd_iommu_set_invalidate_ctx_cb(kfd->adev->pdev, NULL);
+	amd_iommu_set_invalid_ppr_cb(kfd->adev->pdev, NULL);
+	amd_iommu_free_device(kfd->adev->pdev);
 }
 
 /** kfd_iommu_resume - Restore IOMMU after resume
@@ -302,20 +303,20 @@ int kfd_iommu_resume(struct kfd_dev *kfd)
 
 	pasid_limit = kfd_get_pasid_limit();
 
-	err = amd_iommu_init_device(kfd->pdev, pasid_limit);
+	err = amd_iommu_init_device(kfd->adev->pdev, pasid_limit);
 	if (err)
 		return -ENXIO;
 
-	amd_iommu_set_invalidate_ctx_cb(kfd->pdev,
+	amd_iommu_set_invalidate_ctx_cb(kfd->adev->pdev,
 					iommu_pasid_shutdown_callback);
-	amd_iommu_set_invalid_ppr_cb(kfd->pdev,
+	amd_iommu_set_invalid_ppr_cb(kfd->adev->pdev,
 				     iommu_invalid_ppr_cb);
 
 	err = kfd_bind_processes_to_device(kfd);
 	if (err) {
-		amd_iommu_set_invalidate_ctx_cb(kfd->pdev, NULL);
-		amd_iommu_set_invalid_ppr_cb(kfd->pdev, NULL);
-		amd_iommu_free_device(kfd->pdev);
+		amd_iommu_set_invalidate_ctx_cb(kfd->adev->pdev, NULL);
+		amd_iommu_set_invalid_ppr_cb(kfd->adev->pdev, NULL);
+		amd_iommu_free_device(kfd->adev->pdev);
 		return err;
 	}
 
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
index 22b077ac9a19..10048ce16aea 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_migrate.c
@@ -28,7 +28,6 @@
 #include "amdgpu_sync.h"
 #include "amdgpu_object.h"
 #include "amdgpu_vm.h"
-#include "amdgpu_mn.h"
 #include "amdgpu_res_cursor.h"
 #include "kfd_priv.h"
 #include "kfd_svm.h"
@@ -65,8 +64,11 @@ svm_migrate_gart_map(struct amdgpu_ring *ring, uint64_t npages,
 	num_dw = ALIGN(adev->mman.buffer_funcs->copy_num_dw, 8);
 	num_bytes = npages * 8;
 
-	r = amdgpu_job_alloc_with_ib(adev, num_dw * 4 + num_bytes,
-				     AMDGPU_IB_POOL_DELAYED, &job);
+	r = amdgpu_job_alloc_with_ib(adev, &adev->mman.entity,
+				     AMDGPU_FENCE_OWNER_UNDEFINED,
+				     num_dw * 4 + num_bytes,
+				     AMDGPU_IB_POOL_DELAYED,
+				     &job);
 	if (r)
 		return r;
 
@@ -89,18 +91,10 @@ svm_migrate_gart_map(struct amdgpu_ring *ring, uint64_t npages,
 	cpu_addr = &job->ibs[0].ptr[num_dw];
 
 	amdgpu_gart_map(adev, 0, npages, addr, pte_flags, cpu_addr);
-	r = amdgpu_job_submit(job, &adev->mman.entity,
-			      AMDGPU_FENCE_OWNER_UNDEFINED, &fence);
-	if (r)
-		goto error_free;
-
+	fence = amdgpu_job_submit(job);
 	dma_fence_put(fence);
 
 	return r;
-
-error_free:
-	amdgpu_job_free(job);
-	return r;
 }
 
 /**
@@ -529,8 +523,8 @@ svm_migrate_ram_to_vram(struct svm_range *prange, uint32_t best_loc,
 	for (addr = start; addr < end;) {
 		unsigned long next;
 
-		vma = find_vma(mm, addr);
-		if (!vma || addr < vma->vm_start)
+		vma = vma_lookup(mm, addr);
+		if (!vma)
 			break;
 
 		next = min(vma->vm_end, end);
@@ -798,8 +792,8 @@ int svm_migrate_vram_to_ram(struct svm_range *prange, struct mm_struct *mm,
 	for (addr = start; addr < end;) {
 		unsigned long next;
 
-		vma = find_vma(mm, addr);
-		if (!vma || addr < vma->vm_start) {
+		vma = vma_lookup(mm, addr);
+		if (!vma) {
 			pr_debug("failed to find vma for prange %p\n", prange);
 			r = -EFAULT;
 			break;
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_diq.h b/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_diq.h
deleted file mode 100644
index f9cd28690151..000000000000
--- a/drivers/gpu/drm/amd/amdkfd/kfd_pm4_headers_diq.h
+++ /dev/null
@@ -1,291 +0,0 @@
-/* SPDX-License-Identifier: GPL-2.0 OR MIT */
-/*
- * Copyright 2014-2022 Advanced Micro Devices, Inc.
- *
- * Permission is hereby granted, free of charge, to any person obtaining a
- * copy of this software and associated documentation files (the "Software"),
- * to deal in the Software without restriction, including without limitation
- * the rights to use, copy, modify, merge, publish, distribute, sublicense,
- * and/or sell copies of the Software, and to permit persons to whom the
- * Software is furnished to do so, subject to the following conditions:
- *
- * The above copyright notice and this permission notice shall be included in
- * all copies or substantial portions of the Software.
- *
- * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
- * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
- * FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT.  IN NO EVENT SHALL
- * THE COPYRIGHT HOLDER(S) OR AUTHOR(S) BE LIABLE FOR ANY CLAIM, DAMAGES OR
- * OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE,
- * ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR
- * OTHER DEALINGS IN THE SOFTWARE.
- *
- */
-
-#ifndef KFD_PM4_HEADERS_DIQ_H_
-#define KFD_PM4_HEADERS_DIQ_H_
-
-/*--------------------_INDIRECT_BUFFER-------------------- */
-
-#ifndef _PM4__INDIRECT_BUFFER_DEFINED
-#define _PM4__INDIRECT_BUFFER_DEFINED
-enum _INDIRECT_BUFFER_cache_policy_enum {
-	cache_policy___indirect_buffer__lru = 0,
-	cache_policy___indirect_buffer__stream = 1,
-	cache_policy___indirect_buffer__bypass = 2
-};
-
-enum {
-	IT_INDIRECT_BUFFER_PASID = 0x5C
-};
-
-struct pm4__indirect_buffer_pasid {
-	union {
-		union PM4_MES_TYPE_3_HEADER header;	/* header */
-		unsigned int ordinal1;
-	};
-
-	union {
-		struct {
-			unsigned int reserved1:2;
-			unsigned int ib_base_lo:30;
-		} bitfields2;
-		unsigned int ordinal2;
-	};
-
-	union {
-		struct {
-			unsigned int ib_base_hi:16;
-			unsigned int reserved2:16;
-		} bitfields3;
-		unsigned int ordinal3;
-	};
-
-	union {
-		unsigned int control;
-		unsigned int ordinal4;
-	};
-
-	union {
-		struct {
-			unsigned int pasid:10;
-			unsigned int reserved4:22;
-		} bitfields5;
-		unsigned int ordinal5;
-	};
-
-};
-
-#endif
-
-/*--------------------_RELEASE_MEM-------------------- */
-
-#ifndef _PM4__RELEASE_MEM_DEFINED
-#define _PM4__RELEASE_MEM_DEFINED
-enum _RELEASE_MEM_event_index_enum {
-	event_index___release_mem__end_of_pipe = 5,
-	event_index___release_mem__shader_done = 6
-};
-
-enum _RELEASE_MEM_cache_policy_enum {
-	cache_policy___release_mem__lru = 0,
-	cache_policy___release_mem__stream = 1,
-	cache_policy___release_mem__bypass = 2
-};
-
-enum _RELEASE_MEM_dst_sel_enum {
-	dst_sel___release_mem__memory_controller = 0,
-	dst_sel___release_mem__tc_l2 = 1,
-	dst_sel___release_mem__queue_write_pointer_register = 2,
-	dst_sel___release_mem__queue_write_pointer_poll_mask_bit = 3
-};
-
-enum _RELEASE_MEM_int_sel_enum {
-	int_sel___release_mem__none = 0,
-	int_sel___release_mem__send_interrupt_only = 1,
-	int_sel___release_mem__send_interrupt_after_write_confirm = 2,
-	int_sel___release_mem__send_data_after_write_confirm = 3
-};
-
-enum _RELEASE_MEM_data_sel_enum {
-	data_sel___release_mem__none = 0,
-	data_sel___release_mem__send_32_bit_low = 1,
-	data_sel___release_mem__send_64_bit_data = 2,
-	data_sel___release_mem__send_gpu_clock_counter = 3,
-	data_sel___release_mem__send_cp_perfcounter_hi_lo = 4,
-	data_sel___release_mem__store_gds_data_to_memory = 5
-};
-
-struct pm4__release_mem {
-	union {
-		union PM4_MES_TYPE_3_HEADER header;	/*header */
-		unsigned int ordinal1;
-	};
-
-	union {
-		struct {
-			unsigned int event_type:6;
-			unsigned int reserved1:2;
-			enum _RELEASE_MEM_event_index_enum event_index:4;
-			unsigned int tcl1_vol_action_ena:1;
-			unsigned int tc_vol_action_ena:1;
-			unsigned int reserved2:1;
-			unsigned int tc_wb_action_ena:1;
-			unsigned int tcl1_action_ena:1;
-			unsigned int tc_action_ena:1;
-			unsigned int reserved3:6;
-			unsigned int atc:1;
-			enum _RELEASE_MEM_cache_policy_enum cache_policy:2;
-			unsigned int reserved4:5;
-		} bitfields2;
-		unsigned int ordinal2;
-	};
-
-	union {
-		struct {
-			unsigned int reserved5:16;
-			enum _RELEASE_MEM_dst_sel_enum dst_sel:2;
-			unsigned int reserved6:6;
-			enum _RELEASE_MEM_int_sel_enum int_sel:3;
-			unsigned int reserved7:2;
-			enum _RELEASE_MEM_data_sel_enum data_sel:3;
-		} bitfields3;
-		unsigned int ordinal3;
-	};
-
-	union {
-		struct {
-			unsigned int reserved8:2;
-			unsigned int address_lo_32b:30;
-		} bitfields4;
-		struct {
-			unsigned int reserved9:3;
-			unsigned int address_lo_64b:29;
-		} bitfields5;
-		unsigned int ordinal4;
-	};
-
-	unsigned int address_hi;
-
-	unsigned int data_lo;
-
-	unsigned int data_hi;
-
-};
-#endif
-
-
-/*--------------------_SET_CONFIG_REG-------------------- */
-
-#ifndef _PM4__SET_CONFIG_REG_DEFINED
-#define _PM4__SET_CONFIG_REG_DEFINED
-
-struct pm4__set_config_reg {
-	union {
-		union PM4_MES_TYPE_3_HEADER header;	/*header */
-		unsigned int ordinal1;
-	};
-
-	union {
-		struct {
-			unsigned int reg_offset:16;
-			unsigned int reserved1:7;
-			unsigned int vmid_shift:5;
-			unsigned int insert_vmid:1;
-			unsigned int reserved2:3;
-		} bitfields2;
-		unsigned int ordinal2;
-	};
-
-	unsigned int reg_data[1];	/*1..N of these fields */
-
-};
-#endif
-
-/*--------------------_WAIT_REG_MEM-------------------- */
-
-#ifndef _PM4__WAIT_REG_MEM_DEFINED
-#define _PM4__WAIT_REG_MEM_DEFINED
-enum _WAIT_REG_MEM_function_enum {
-	function___wait_reg_mem__always_pass = 0,
-	function___wait_reg_mem__less_than_ref_value = 1,
-	function___wait_reg_mem__less_than_equal_to_the_ref_value = 2,
-	function___wait_reg_mem__equal_to_the_reference_value = 3,
-	function___wait_reg_mem__not_equal_reference_value = 4,
-	function___wait_reg_mem__greater_than_or_equal_reference_value = 5,
-	function___wait_reg_mem__greater_than_reference_value = 6,
-	function___wait_reg_mem__reserved = 7
-};
-
-enum _WAIT_REG_MEM_mem_space_enum {
-	mem_space___wait_reg_mem__register_space = 0,
-	mem_space___wait_reg_mem__memory_space = 1
-};
-
-enum _WAIT_REG_MEM_operation_enum {
-	operation___wait_reg_mem__wait_reg_mem = 0,
-	operation___wait_reg_mem__wr_wait_wr_reg = 1
-};
-
-struct pm4__wait_reg_mem {
-	union {
-		union PM4_MES_TYPE_3_HEADER header;	/*header */
-		unsigned int ordinal1;
-	};
-
-	union {
-		struct {
-			enum _WAIT_REG_MEM_function_enum function:3;
-			unsigned int reserved1:1;
-			enum _WAIT_REG_MEM_mem_space_enum mem_space:2;
-			enum _WAIT_REG_MEM_operation_enum operation:2;
-			unsigned int reserved2:24;
-		} bitfields2;
-		unsigned int ordinal2;
-	};
-
-	union {
-		struct {
-			unsigned int reserved3:2;
-			unsigned int memory_poll_addr_lo:30;
-		} bitfields3;
-		struct {
-			unsigned int register_poll_addr:16;
-			unsigned int reserved4:16;
-		} bitfields4;
-		struct {
-			unsigned int register_write_addr:16;
-			unsigned int reserved5:16;
-		} bitfields5;
-		unsigned int ordinal3;
-	};
-
-	union {
-		struct {
-			unsigned int poll_address_hi:16;
-			unsigned int reserved6:16;
-		} bitfields6;
-		struct {
-			unsigned int register_write_addr:16;
-			unsigned int reserved7:16;
-		} bitfields7;
-		unsigned int ordinal4;
-	};
-
-	unsigned int reference;
-
-	unsigned int mask;
-
-	union {
-		struct {
-			unsigned int poll_interval:16;
-			unsigned int reserved8:16;
-		} bitfields8;
-		unsigned int ordinal7;
-	};
-
-};
-#endif
-
-
-#endif /* KFD_PM4_HEADERS_DIQ_H_ */
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
index bf610e3b683b..552c3ac85a13 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_priv.h
@@ -254,8 +254,6 @@ struct kfd_dev {
 	struct amdgpu_device *adev;
 
 	struct kfd_device_info device_info;
-	struct pci_dev *pdev;
-	struct drm_device *ddev;
 
 	unsigned int id;		/* topology stub index */
 
@@ -1365,7 +1363,7 @@ void kfd_dec_compute_active(struct kfd_dev *dev);
 static inline int kfd_devcgroup_check_permission(struct kfd_dev *kfd)
 {
 #if defined(CONFIG_CGROUP_DEVICE) || defined(CONFIG_CGROUP_BPF)
-	struct drm_device *ddev = kfd->ddev;
+	struct drm_device *ddev = adev_to_drm(kfd->adev);
 
 	return devcgroup_check_permission(DEVCG_DEV_CHAR, DRM_MAJOR,
 					  ddev->render->index,
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_process.c b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
index 951b63677248..a26257171ab7 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_process.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_process.c
@@ -1050,8 +1050,8 @@ static void kfd_process_destroy_pdds(struct kfd_process *p)
 		 * for auto suspend
 		 */
 		if (pdd->runtime_inuse) {
-			pm_runtime_mark_last_busy(pdd->dev->ddev->dev);
-			pm_runtime_put_autosuspend(pdd->dev->ddev->dev);
+			pm_runtime_mark_last_busy(adev_to_drm(pdd->dev->adev)->dev);
+			pm_runtime_put_autosuspend(adev_to_drm(pdd->dev->adev)->dev);
 			pdd->runtime_inuse = false;
 		}
 
@@ -1633,9 +1633,9 @@ struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev,
 	 * pdd is destroyed.
 	 */
 	if (!pdd->runtime_inuse) {
-		err = pm_runtime_get_sync(dev->ddev->dev);
+		err = pm_runtime_get_sync(adev_to_drm(dev->adev)->dev);
 		if (err < 0) {
-			pm_runtime_put_autosuspend(dev->ddev->dev);
+			pm_runtime_put_autosuspend(adev_to_drm(dev->adev)->dev);
 			return ERR_PTR(err);
 		}
 	}
@@ -1655,8 +1655,8 @@ struct kfd_process_device *kfd_bind_process_to_device(struct kfd_dev *dev,
 out:
 	/* balance runpm reference count and exit with error */
 	if (!pdd->runtime_inuse) {
-		pm_runtime_mark_last_busy(dev->ddev->dev);
-		pm_runtime_put_autosuspend(dev->ddev->dev);
+		pm_runtime_mark_last_busy(adev_to_drm(dev->adev)->dev);
+		pm_runtime_put_autosuspend(adev_to_drm(dev->adev)->dev);
 	}
 
 	return ERR_PTR(err);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
index 64fdf63093a0..814f99888ab1 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_svm.c
@@ -26,7 +26,7 @@
 #include "amdgpu_sync.h"
 #include "amdgpu_object.h"
 #include "amdgpu_vm.h"
-#include "amdgpu_mn.h"
+#include "amdgpu_hmm.h"
 #include "amdgpu.h"
 #include "amdgpu_xgmi.h"
 #include "kfd_priv.h"
@@ -259,7 +259,7 @@ void svm_range_free_dma_mappings(struct svm_range *prange)
 			pr_debug("failed to find device idx %d\n", gpuidx);
 			continue;
 		}
-		dev = &pdd->dev->pdev->dev;
+		dev = &pdd->dev->adev->pdev->dev;
 		svm_range_dma_unmap(dev, dma_addr, 0, prange->npages);
 		kvfree(dma_addr);
 		prange->dma_addr[gpuidx] = NULL;
@@ -1586,8 +1586,8 @@ static int svm_range_validate_and_map(struct mm_struct *mm,
 		unsigned long npages;
 		bool readonly;
 
-		vma = find_vma(mm, addr);
-		if (!vma || addr < vma->vm_start) {
+		vma = vma_lookup(mm, addr);
+		if (!vma) {
 			r = -EFAULT;
 			goto unreserve_out;
 		}
@@ -1596,9 +1596,9 @@ static int svm_range_validate_and_map(struct mm_struct *mm,
 		next = min(vma->vm_end, end);
 		npages = (next - addr) >> PAGE_SHIFT;
 		WRITE_ONCE(p->svms.faulting_task, current);
-		r = amdgpu_hmm_range_get_pages(&prange->notifier, mm, NULL,
-					       addr, npages, &hmm_range,
-					       readonly, true, owner);
+		r = amdgpu_hmm_range_get_pages(&prange->notifier, addr, npages,
+					       readonly, owner, NULL,
+					       &hmm_range);
 		WRITE_ONCE(p->svms.faulting_task, NULL);
 		if (r) {
 			pr_debug("failed %d to get svm range pages\n", r);
@@ -2542,8 +2542,8 @@ svm_range_get_range_boundaries(struct kfd_process *p, int64_t addr,
 	struct interval_tree_node *node;
 	unsigned long start_limit, end_limit;
 
-	vma = find_vma(p->mm, addr << PAGE_SHIFT);
-	if (!vma || (addr << PAGE_SHIFT) < vma->vm_start) {
+	vma = vma_lookup(p->mm, addr << PAGE_SHIFT);
+	if (!vma) {
 		pr_debug("VMA does not exist in address [0x%llx]\n", addr);
 		return -EFAULT;
 	}
@@ -2871,8 +2871,8 @@ retry_write_locked:
 	/* __do_munmap removed VMA, return success as we are handling stale
 	 * retry fault.
 	 */
-	vma = find_vma(mm, addr << PAGE_SHIFT);
-	if (!vma || (addr << PAGE_SHIFT) < vma->vm_start) {
+	vma = vma_lookup(mm, addr << PAGE_SHIFT);
+	if (!vma) {
 		pr_debug("address 0x%llx VMA is removed\n", addr);
 		r = 0;
 		goto out_unlock_range;
@@ -3152,9 +3152,8 @@ svm_range_is_valid(struct kfd_process *p, uint64_t start, uint64_t size)
 	start <<= PAGE_SHIFT;
 	end = start + (size << PAGE_SHIFT);
 	do {
-		vma = find_vma(p->mm, start);
-		if (!vma || start < vma->vm_start ||
-		    (vma->vm_flags & device_vma))
+		vma = vma_lookup(p->mm, start);
+		if (!vma || (vma->vm_flags & device_vma))
 			return -EFAULT;
 		start = min(end, vma->vm_end);
 	} while (start < end);
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
index 3f0a4a415907..bceb1a5b2518 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.c
@@ -115,7 +115,7 @@ struct kfd_dev *kfd_device_by_pci_dev(const struct pci_dev *pdev)
 	down_read(&topology_lock);
 
 	list_for_each_entry(top_dev, &topology_device_list, list)
-		if (top_dev->gpu && top_dev->gpu->pdev == pdev) {
+		if (top_dev->gpu && top_dev->gpu->adev->pdev == pdev) {
 			device = top_dev->gpu;
 			break;
 		}
@@ -364,7 +364,6 @@ static ssize_t kfd_cache_show(struct kobject *kobj, struct attribute *attr,
 
 	/* Making sure that the buffer is an empty string */
 	buffer[0] = 0;
-
 	cache = container_of(attr, struct kfd_cache_properties, attr);
 	if (cache->gpu && kfd_devcgroup_check_permission(cache->gpu))
 		return -EPERM;
@@ -379,12 +378,13 @@ static ssize_t kfd_cache_show(struct kobject *kobj, struct attribute *attr,
 	sysfs_show_32bit_prop(buffer, offs, "association", cache->cache_assoc);
 	sysfs_show_32bit_prop(buffer, offs, "latency", cache->cache_latency);
 	sysfs_show_32bit_prop(buffer, offs, "type", cache->cache_type);
+
 	offs += snprintf(buffer+offs, PAGE_SIZE-offs, "sibling_map ");
-	for (i = 0; i < CRAT_SIBLINGMAP_SIZE; i++)
+	for (i = 0; i < cache->sibling_map_size; i++)
 		for (j = 0; j < sizeof(cache->sibling_map[0])*8; j++)
 			/* Check each bit */
 			offs += snprintf(buffer+offs, PAGE_SIZE-offs, "%d,",
-					 (cache->sibling_map[i] >> j) & 1);
+						(cache->sibling_map[i] >> j) & 1);
 
 	/* Replace the last "," with end of line */
 	buffer[offs-1] = '\n';
@@ -1169,13 +1169,12 @@ static uint32_t kfd_generate_gpu_id(struct kfd_dev *gpu)
 
 	local_mem_size = gpu->local_mem_info.local_mem_size_private +
 			gpu->local_mem_info.local_mem_size_public;
-
-	buf[0] = gpu->pdev->devfn;
-	buf[1] = gpu->pdev->subsystem_vendor |
-		(gpu->pdev->subsystem_device << 16);
-	buf[2] = pci_domain_nr(gpu->pdev->bus);
-	buf[3] = gpu->pdev->device;
-	buf[4] = gpu->pdev->bus->number;
+	buf[0] = gpu->adev->pdev->devfn;
+	buf[1] = gpu->adev->pdev->subsystem_vendor |
+		(gpu->adev->pdev->subsystem_device << 16);
+	buf[2] = pci_domain_nr(gpu->adev->pdev->bus);
+	buf[3] = gpu->adev->pdev->device;
+	buf[4] = gpu->adev->pdev->bus->number;
 	buf[5] = lower_32_bits(local_mem_size);
 	buf[6] = upper_32_bits(local_mem_size);
 
@@ -1198,7 +1197,6 @@ static struct kfd_topology_device *kfd_assign_gpu(struct kfd_dev *gpu)
 	struct kfd_iolink_properties *iolink;
 	struct kfd_iolink_properties *p2plink;
 
-	down_write(&topology_lock);
 	list_for_each_entry(dev, &topology_device_list, list) {
 		/* Discrete GPUs need their own topology device list
 		 * entries. Don't assign them to CPU/APU nodes.
@@ -1222,7 +1220,6 @@ static struct kfd_topology_device *kfd_assign_gpu(struct kfd_dev *gpu)
 			break;
 		}
 	}
-	up_write(&topology_lock);
 	return out_dev;
 }
 
@@ -1269,7 +1266,7 @@ static void kfd_set_iolink_no_atomics(struct kfd_topology_device *dev,
 	if (target_gpu_dev) {
 		uint32_t cap;
 
-		pcie_capability_read_dword(target_gpu_dev->gpu->pdev,
+		pcie_capability_read_dword(target_gpu_dev->gpu->adev->pdev,
 				PCI_EXP_DEVCAP2, &cap);
 
 		if (!(cap & (PCI_EXP_DEVCAP2_ATOMIC_COMP32 |
@@ -1593,21 +1590,290 @@ out:
 	return ret;
 }
 
+
+/* Helper function. See kfd_fill_gpu_cache_info for parameter description */
+static int fill_in_l1_pcache(struct kfd_cache_properties **props_ext,
+				struct kfd_gpu_cache_info *pcache_info,
+				struct kfd_cu_info *cu_info,
+				int cu_bitmask,
+				int cache_type, unsigned int cu_processor_id,
+				int cu_block)
+{
+	unsigned int cu_sibling_map_mask;
+	int first_active_cu;
+	struct kfd_cache_properties *pcache = NULL;
+
+	cu_sibling_map_mask = cu_bitmask;
+	cu_sibling_map_mask >>= cu_block;
+	cu_sibling_map_mask &= ((1 << pcache_info[cache_type].num_cu_shared) - 1);
+	first_active_cu = ffs(cu_sibling_map_mask);
+
+	/* CU could be inactive. In case of shared cache find the first active
+	 * CU. and incase of non-shared cache check if the CU is inactive. If
+	 * inactive active skip it
+	 */
+	if (first_active_cu) {
+		pcache = kfd_alloc_struct(pcache);
+		if (!pcache)
+			return -ENOMEM;
+
+		memset(pcache, 0, sizeof(struct kfd_cache_properties));
+		pcache->processor_id_low = cu_processor_id + (first_active_cu - 1);
+		pcache->cache_level = pcache_info[cache_type].cache_level;
+		pcache->cache_size = pcache_info[cache_type].cache_size;
+
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_DATA_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_DATA;
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_INST_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_INSTRUCTION;
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_CPU_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_CPU;
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_SIMD_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_HSACU;
+
+		/* Sibling map is w.r.t processor_id_low, so shift out
+		 * inactive CU
+		 */
+		cu_sibling_map_mask =
+			cu_sibling_map_mask >> (first_active_cu - 1);
+
+		pcache->sibling_map[0] = (uint8_t)(cu_sibling_map_mask & 0xFF);
+		pcache->sibling_map[1] =
+				(uint8_t)((cu_sibling_map_mask >> 8) & 0xFF);
+		pcache->sibling_map[2] =
+				(uint8_t)((cu_sibling_map_mask >> 16) & 0xFF);
+		pcache->sibling_map[3] =
+				(uint8_t)((cu_sibling_map_mask >> 24) & 0xFF);
+
+		pcache->sibling_map_size = 4;
+		*props_ext = pcache;
+
+		return 0;
+	}
+	return 1;
+}
+
+/* Helper function. See kfd_fill_gpu_cache_info for parameter description */
+static int fill_in_l2_l3_pcache(struct kfd_cache_properties **props_ext,
+				struct kfd_gpu_cache_info *pcache_info,
+				struct kfd_cu_info *cu_info,
+				int cache_type, unsigned int cu_processor_id)
+{
+	unsigned int cu_sibling_map_mask;
+	int first_active_cu;
+	int i, j, k;
+	struct kfd_cache_properties *pcache = NULL;
+
+	cu_sibling_map_mask = cu_info->cu_bitmap[0][0];
+	cu_sibling_map_mask &=
+		((1 << pcache_info[cache_type].num_cu_shared) - 1);
+	first_active_cu = ffs(cu_sibling_map_mask);
+
+	/* CU could be inactive. In case of shared cache find the first active
+	 * CU. and incase of non-shared cache check if the CU is inactive. If
+	 * inactive active skip it
+	 */
+	if (first_active_cu) {
+		pcache = kfd_alloc_struct(pcache);
+		if (!pcache)
+			return -ENOMEM;
+
+		memset(pcache, 0, sizeof(struct kfd_cache_properties));
+		pcache->processor_id_low = cu_processor_id
+					+ (first_active_cu - 1);
+		pcache->cache_level = pcache_info[cache_type].cache_level;
+		pcache->cache_size = pcache_info[cache_type].cache_size;
+
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_DATA_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_DATA;
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_INST_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_INSTRUCTION;
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_CPU_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_CPU;
+		if (pcache_info[cache_type].flags & CRAT_CACHE_FLAGS_SIMD_CACHE)
+			pcache->cache_type |= HSA_CACHE_TYPE_HSACU;
+
+		/* Sibling map is w.r.t processor_id_low, so shift out
+		 * inactive CU
+		 */
+		cu_sibling_map_mask = cu_sibling_map_mask >> (first_active_cu - 1);
+		k = 0;
+
+		for (i = 0; i < cu_info->num_shader_engines; i++) {
+			for (j = 0; j < cu_info->num_shader_arrays_per_engine; j++) {
+				pcache->sibling_map[k] = (uint8_t)(cu_sibling_map_mask & 0xFF);
+				pcache->sibling_map[k+1] = (uint8_t)((cu_sibling_map_mask >> 8) & 0xFF);
+				pcache->sibling_map[k+2] = (uint8_t)((cu_sibling_map_mask >> 16) & 0xFF);
+				pcache->sibling_map[k+3] = (uint8_t)((cu_sibling_map_mask >> 24) & 0xFF);
+				k += 4;
+
+				cu_sibling_map_mask = cu_info->cu_bitmap[i % 4][j + i / 4];
+				cu_sibling_map_mask &= ((1 << pcache_info[cache_type].num_cu_shared) - 1);
+			}
+		}
+		pcache->sibling_map_size = k;
+		*props_ext = pcache;
+		return 0;
+	}
+	return 1;
+}
+
+#define KFD_MAX_CACHE_TYPES 6
+
+/* kfd_fill_cache_non_crat_info - Fill GPU cache info using kfd_gpu_cache_info
+ * tables
+ */
+static void kfd_fill_cache_non_crat_info(struct kfd_topology_device *dev, struct kfd_dev *kdev)
+{
+	struct kfd_gpu_cache_info *pcache_info = NULL;
+	int i, j, k;
+	int ct = 0;
+	unsigned int cu_processor_id;
+	int ret;
+	unsigned int num_cu_shared;
+	struct kfd_cu_info cu_info;
+	struct kfd_cu_info *pcu_info;
+	int gpu_processor_id;
+	struct kfd_cache_properties *props_ext;
+	int num_of_entries = 0;
+	int num_of_cache_types = 0;
+	struct kfd_gpu_cache_info cache_info[KFD_MAX_CACHE_TYPES];
+
+	amdgpu_amdkfd_get_cu_info(kdev->adev, &cu_info);
+	pcu_info = &cu_info;
+
+	gpu_processor_id = dev->node_props.simd_id_base;
+
+	pcache_info = cache_info;
+	num_of_cache_types = kfd_get_gpu_cache_info(kdev, &pcache_info);
+	if (!num_of_cache_types) {
+		pr_warn("no cache info found\n");
+		return;
+	}
+
+	/* For each type of cache listed in the kfd_gpu_cache_info table,
+	 * go through all available Compute Units.
+	 * The [i,j,k] loop will
+	 *		if kfd_gpu_cache_info.num_cu_shared = 1
+	 *			will parse through all available CU
+	 *		If (kfd_gpu_cache_info.num_cu_shared != 1)
+	 *			then it will consider only one CU from
+	 *			the shared unit
+	 */
+	for (ct = 0; ct < num_of_cache_types; ct++) {
+		cu_processor_id = gpu_processor_id;
+		if (pcache_info[ct].cache_level == 1) {
+			for (i = 0; i < pcu_info->num_shader_engines; i++) {
+				for (j = 0; j < pcu_info->num_shader_arrays_per_engine; j++) {
+					for (k = 0; k < pcu_info->num_cu_per_sh; k += pcache_info[ct].num_cu_shared) {
+
+						ret = fill_in_l1_pcache(&props_ext, pcache_info, pcu_info,
+										pcu_info->cu_bitmap[i % 4][j + i / 4], ct,
+										cu_processor_id, k);
+
+						if (ret < 0)
+							break;
+
+						if (!ret) {
+							num_of_entries++;
+							list_add_tail(&props_ext->list, &dev->cache_props);
+						}
+
+						/* Move to next CU block */
+						num_cu_shared = ((k + pcache_info[ct].num_cu_shared) <=
+							pcu_info->num_cu_per_sh) ?
+							pcache_info[ct].num_cu_shared :
+							(pcu_info->num_cu_per_sh - k);
+						cu_processor_id += num_cu_shared;
+					}
+				}
+			}
+		} else {
+			ret = fill_in_l2_l3_pcache(&props_ext, pcache_info,
+								pcu_info, ct, cu_processor_id);
+
+			if (ret < 0)
+				break;
+
+			if (!ret) {
+				num_of_entries++;
+				list_add_tail(&props_ext->list, &dev->cache_props);
+			}
+		}
+	}
+	dev->node_props.caches_count += num_of_entries;
+	pr_debug("Added [%d] GPU cache entries\n", num_of_entries);
+}
+
+static int kfd_topology_add_device_locked(struct kfd_dev *gpu, uint32_t gpu_id,
+					  struct kfd_topology_device **dev)
+{
+	int proximity_domain = ++topology_crat_proximity_domain;
+	struct list_head temp_topology_device_list;
+	void *crat_image = NULL;
+	size_t image_size = 0;
+	int res;
+
+	res = kfd_create_crat_image_virtual(&crat_image, &image_size,
+					    COMPUTE_UNIT_GPU, gpu,
+					    proximity_domain);
+	if (res) {
+		pr_err("Error creating VCRAT for GPU (ID: 0x%x)\n",
+		       gpu_id);
+		topology_crat_proximity_domain--;
+		goto err;
+	}
+
+	INIT_LIST_HEAD(&temp_topology_device_list);
+
+	res = kfd_parse_crat_table(crat_image,
+				   &temp_topology_device_list,
+				   proximity_domain);
+	if (res) {
+		pr_err("Error parsing VCRAT for GPU (ID: 0x%x)\n",
+		       gpu_id);
+		topology_crat_proximity_domain--;
+		goto err;
+	}
+
+	kfd_topology_update_device_list(&temp_topology_device_list,
+					&topology_device_list);
+
+	*dev = kfd_assign_gpu(gpu);
+	if (WARN_ON(!*dev)) {
+		res = -ENODEV;
+		goto err;
+	}
+
+	/* Fill the cache affinity information here for the GPUs
+	 * using VCRAT
+	 */
+	kfd_fill_cache_non_crat_info(*dev, gpu);
+
+	/* Update the SYSFS tree, since we added another topology
+	 * device
+	 */
+	res = kfd_topology_update_sysfs();
+	if (!res)
+		sys_props.generation_count++;
+	else
+		pr_err("Failed to update GPU (ID: 0x%x) to sysfs topology. res=%d\n",
+		       gpu_id, res);
+
+err:
+	kfd_destroy_crat_image(crat_image);
+	return res;
+}
+
 int kfd_topology_add_device(struct kfd_dev *gpu)
 {
 	uint32_t gpu_id;
 	struct kfd_topology_device *dev;
 	struct kfd_cu_info cu_info;
 	int res = 0;
-	struct list_head temp_topology_device_list;
-	void *crat_image = NULL;
-	size_t image_size = 0;
-	int proximity_domain;
 	int i;
 	const char *asic_name = amdgpu_asic_name[gpu->adev->asic_type];
 
-	INIT_LIST_HEAD(&temp_topology_device_list);
-
 	gpu_id = kfd_generate_gpu_id(gpu);
 	pr_debug("Adding new GPU (ID: 0x%x) to topology\n", gpu_id);
 
@@ -1617,50 +1883,13 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 	 * CRAT to create a new topology device. Once created assign the gpu to
 	 * that topology device
 	 */
+	down_write(&topology_lock);
 	dev = kfd_assign_gpu(gpu);
-	if (!dev) {
-		down_write(&topology_lock);
-		proximity_domain = ++topology_crat_proximity_domain;
-
-		res = kfd_create_crat_image_virtual(&crat_image, &image_size,
-						    COMPUTE_UNIT_GPU, gpu,
-						    proximity_domain);
-		if (res) {
-			pr_err("Error creating VCRAT for GPU (ID: 0x%x)\n",
-			       gpu_id);
-			topology_crat_proximity_domain--;
-			return res;
-		}
-		res = kfd_parse_crat_table(crat_image,
-					   &temp_topology_device_list,
-					   proximity_domain);
-		if (res) {
-			pr_err("Error parsing VCRAT for GPU (ID: 0x%x)\n",
-			       gpu_id);
-			topology_crat_proximity_domain--;
-			goto err;
-		}
-
-		kfd_topology_update_device_list(&temp_topology_device_list,
-			&topology_device_list);
-
-		/* Update the SYSFS tree, since we added another topology
-		 * device
-		 */
-		res = kfd_topology_update_sysfs();
-		up_write(&topology_lock);
-
-		if (!res)
-			sys_props.generation_count++;
-		else
-			pr_err("Failed to update GPU (ID: 0x%x) to sysfs topology. res=%d\n",
-						gpu_id, res);
-		dev = kfd_assign_gpu(gpu);
-		if (WARN_ON(!dev)) {
-			res = -ENODEV;
-			goto err;
-		}
-	}
+	if (!dev)
+		res = kfd_topology_add_device_locked(gpu, gpu_id, &dev);
+	up_write(&topology_lock);
+	if (res)
+		return res;
 
 	dev->gpu_id = gpu_id;
 	gpu->id = gpu_id;
@@ -1688,13 +1917,13 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 		cu_info.num_shader_arrays_per_engine;
 
 	dev->node_props.gfx_target_version = gpu->device_info.gfx_target_version;
-	dev->node_props.vendor_id = gpu->pdev->vendor;
-	dev->node_props.device_id = gpu->pdev->device;
+	dev->node_props.vendor_id = gpu->adev->pdev->vendor;
+	dev->node_props.device_id = gpu->adev->pdev->device;
 	dev->node_props.capability |=
 		((dev->gpu->adev->rev_id << HSA_CAP_ASIC_REVISION_SHIFT) &
 			HSA_CAP_ASIC_REVISION_MASK);
-	dev->node_props.location_id = pci_dev_id(gpu->pdev);
-	dev->node_props.domain = pci_domain_nr(gpu->pdev->bus);
+	dev->node_props.location_id = pci_dev_id(gpu->adev->pdev);
+	dev->node_props.domain = pci_domain_nr(gpu->adev->pdev->bus);
 	dev->node_props.max_engine_clk_fcompute =
 		amdgpu_amdkfd_get_max_engine_clock_in_mhz(dev->gpu->adev);
 	dev->node_props.max_engine_clk_ccompute =
@@ -1783,11 +2012,9 @@ int kfd_topology_add_device(struct kfd_dev *gpu)
 
 	kfd_debug_print_topology();
 
-	if (!res)
-		kfd_notify_gpu_change(gpu_id, 1);
-err:
-	kfd_destroy_crat_image(crat_image);
-	return res;
+	kfd_notify_gpu_change(gpu_id, 1);
+
+	return 0;
 }
 
 /**
diff --git a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
index 9f6c949186c1..fca30d00a9bb 100644
--- a/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
+++ b/drivers/gpu/drm/amd/amdkfd/kfd_topology.h
@@ -80,6 +80,8 @@ struct kfd_mem_properties {
 	struct attribute	attr;
 };
 
+#define CACHE_SIBLINGMAP_SIZE 64
+
 struct kfd_cache_properties {
 	struct list_head	list;
 	uint32_t		processor_id_low;
@@ -90,10 +92,11 @@ struct kfd_cache_properties {
 	uint32_t		cache_assoc;
 	uint32_t		cache_latency;
 	uint32_t		cache_type;
-	uint8_t			sibling_map[CRAT_SIBLINGMAP_SIZE];
+	uint8_t			sibling_map[CACHE_SIBLINGMAP_SIZE];
 	struct kfd_dev		*gpu;
 	struct kobject		*kobj;
 	struct attribute	attr;
+	uint32_t		sibling_map_size;
 };
 
 struct kfd_iolink_properties {
@@ -128,7 +131,6 @@ struct kfd_topology_device {
 	uint32_t			proximity_domain;
 	struct kfd_node_properties	node_props;
 	struct list_head		mem_props;
-	uint32_t			cache_count;
 	struct list_head		cache_props;
 	struct list_head		io_link_props;
 	struct list_head		p2p_link_props;