Merge tag 'drm-next-2020-01-30' of git://anongit.freedesktop.org/drm/drm

Pull drm updates from Davbe Airlie:
 "This is the main pull request for graphics for 5.6. Usual selection of
  changes all over.

  I've got one outstanding vmwgfx pull that touches mm so kept it
  separate until after all of this lands. I'll try and get it to you
  soon after this, but it might be early next week (nothing wrong with
  code, just my schedule is messy)

  This also hits a lot of fbdev drivers with some cleanups.

  Other notables:
   - vulkan timeline semaphore support added to syncobjs
   - nouveau turing secureboot/graphics support
   - Displayport MST display stream compression support

  Detailed summary:

  uapi:
   - dma-buf heaps added (and fixed)
   - command line add support for panel oreientation
   - command line allow overriding penguin count

  drm:
   - mipi dsi definition updates
   - lockdep annotations for dma_resv
   - remove dma-buf kmap/kunmap support
   - constify fb_ops in all fbdev drivers
   - MST fix for daisy chained hotplug-
   - CTA-861-G modes with VIC >= 193 added
   - fix drm_panel_of_backlight export
   - LVDS decoder support
   - more device based logging support
   - scanline alighment for dumb buffers
   - MST DSC helpers

  scheduler:
   - documentation fixes
   - job distribution improvements

  panel:
   - Logic PD type 28 panel support
   - Jimax8729d MIPI-DSI
   - igenic JZ4770
   - generic DSI devicetree bindings
   - sony acx424AKP panel
   - Leadtek LTK500HD1829
   - xinpeng XPP055C272
   - AUO B116XAK01
   - GiantPlus GPM940B0
   - BOE NV140FHM-N49
   - Satoz SAT050AT40H12R2
   - Sharp LS020B1DD01D panels.

  ttm:
   - use blocking WW lock

  i915:
   - hw/uapi state separation
   - Lock annotation improvements
   - selftest improvements
   - ICL/TGL DSI VDSC support
   - VBT parsing improvments
   - Display refactoring
   - DSI updates + fixes
   - HDCP 2.2 for CFL
   - CML PCI ID fixes
   - GLK+ fbc fix
   - PSR fixes
   - GEN/GT refactor improvments
   - DP MST fixes
   - switch context id alloc to xarray
   - workaround updates
   - LMEM debugfs support
   - tiled monitor fixes
   - ICL+ clock gating programming removed
   - DP MST disable sequence fixed
   - LMEM discontiguous object maps
   - prefaulting for discontiguous objects
   - use LMEM for dumb buffers if possible
   - add LMEM mmap support

  amdgpu:
   - enable sync object timelines for vulkan
   - MST atomic routines
   - enable MST DSC support
   - add DMCUB display microengine support
   - DC OEM i2c support
   - Renoir DC fixes
   - Initial HDCP 2.x support
   - BACO support for Arcturus
   - Use BACO for runtime PM power save
   - gfxoff on navi10
   - gfx10 golden updates and fixes
   - DCN support on POWER
   - GFXOFF for raven1 refresh
   - MM engine idle handlers cleanup
   - 10bpc EDP panel fixes
   - renoir watermark fixes
   - SR-IOV fixes
   - Arcturus VCN fixes
   - GDDR6 training fixes
   - freesync fixes
   - Pollock support

  amdkfd:
   - unify more codepath with amdgpu
   - use KIQ to setup HIQ rather than MMIO

  radeon:
   - fix vma fault handler race
   - PPC DMA fix
   - register check fixes for r100/r200

  nouveau:
   - mmap_sem vs dma_resv fix
   - rewrite the ACR secure boot code for Turing
   - TU10x graphics engine support (TU11x pending)
   - Page kind mapping for turing
   - 10-bit LUT support
   - GP10B Tegra fixes
   - HD audio regression fix

  hisilicon/hibmc:
   - use generic fbdev code and helpers

  rockchip:
   - dsi/px30 support

  virtio:
   - fb damage support
   - static some functions

  vc4:
   - use dma_resv lock wrappers

  msm:
   - use dma_resv lock wrappers
   - sc7180 display + DSI support
   - a618 support
   - UBWC support improvements

  vmwgfx:
   - updates + new logging uapi

  exynos:
   - enable/disable callback cleanups

  etnaviv:
   - use dma_resv lock wrappers

  atmel-hlcdc:
   - clock fixes

  mediatek:
   - cmdq support
   - non-smooth cursor fixes
   - ctm property support

  sun4i:
   - suspend support
   - A64 mipi dsi support

  rcar-du:
   - Color management module support
   - LVDS encoder dual-link support
   - R8A77980 support

  analogic:
   - add support for an6345

  ast:
   - atomic modeset support
   - primary plane garbage fix

  arcgpu:
   - fixes for fourcc handling

  tegra:
   - minor fixes and improvments

  mcde:
   - vblank support

  meson:
   - OSD1 plane AFBC commit

  gma500:
   - add pageflip support
   - reomve global drm_dev

  komeda:
   - tweak debugfs output
   - d32 support
   - runtime PM suppotr

  udl:
   - use generic shmem helpers
   - cleanup and fixes"

* tag 'drm-next-2020-01-30' of git://anongit.freedesktop.org/drm/drm: (1998 commits)
  drm/nouveau/fb/gp102-: allow module to load even when scrubber binary is missing
  drm/nouveau/acr: return error when registering LSF if ACR not supported
  drm/nouveau/disp/gv100-: not all channel types support reporting error codes
  drm/nouveau/disp/nv50-: prevent oops when no channel method map provided
  drm/nouveau: support synchronous pushbuf submission
  drm/nouveau: signal pending fences when channel has been killed
  drm/nouveau: reject attempts to submit to dead channels
  drm/nouveau: zero vma pointer even if we only unreference it rather than free
  drm/nouveau: Add HD-audio component notifier support
  drm/nouveau: fix build error without CONFIG_IOMMU_API
  drm/nouveau/kms/nv04: remove set but not used variable 'width'
  drm/nouveau/kms/nv50: remove set but not unused variable 'nv_connector'
  drm/nouveau/mmu: fix comptag memory leak
  drm/nouveau/gr/gp10b: Use gp100_grctx and gp100_gr_zbc
  drm/nouveau/pmu/gm20b,gp10b: Fix Falcon bootstrapping
  drm/exynos: Rename Exynos to lowercase
  drm/exynos: change callback names
  drm/mst: Don't do atomic checks over disabled managers
  drm/amdgpu: add the lost mutex_init back
  drm/amd/display: skip opp blank or unblank if test pattern enabled
  ...

1 files changed, 261 insertions, 293 deletions

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
index e1b8d8daeafc..3a1570dafe34 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_psp.c
@@ -158,7 +158,7 @@ psp_cmd_submit_buf(struct psp_context *psp,
 	memcpy(psp->cmd_buf_mem, cmd, sizeof(struct psp_gfx_cmd_resp));
 
 	index = atomic_inc_return(&psp->fence_value);
-	ret = psp_cmd_submit(psp, psp->cmd_buf_mc_addr, fence_mc_addr, index);
+	ret = psp_ring_cmd_submit(psp, psp->cmd_buf_mc_addr, fence_mc_addr, index);
 	if (ret) {
 		atomic_dec(&psp->fence_value);
 		mutex_unlock(&psp->mutex);
@@ -191,9 +191,9 @@ psp_cmd_submit_buf(struct psp_context *psp,
 		if (ucode)
 			DRM_WARN("failed to load ucode id (%d) ",
 				  ucode->ucode_id);
-		DRM_DEBUG_DRIVER("psp command (0x%X) failed and response status is (0x%X)\n",
+		DRM_WARN("psp command (0x%X) failed and response status is (0x%X)\n",
 			 psp->cmd_buf_mem->cmd_id,
-			 psp->cmd_buf_mem->resp.status & GFX_CMD_STATUS_MASK);
+			 psp->cmd_buf_mem->resp.status);
 		if (!timeout) {
 			mutex_unlock(&psp->mutex);
 			return -EINVAL;
@@ -318,35 +318,17 @@ static int psp_tmr_load(struct psp_context *psp)
 	return ret;
 }
 
-static void psp_prep_asd_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-				 uint64_t asd_mc, uint64_t asd_mc_shared,
-				 uint32_t size, uint32_t shared_size)
+static void psp_prep_asd_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+				uint64_t asd_mc, uint32_t size)
 {
 	cmd->cmd_id = GFX_CMD_ID_LOAD_ASD;
 	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(asd_mc);
 	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(asd_mc);
 	cmd->cmd.cmd_load_ta.app_len = size;
 
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(asd_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(asd_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
-}
-
-static int psp_asd_init(struct psp_context *psp)
-{
-	int ret;
-
-	/*
-	 * Allocate 16k memory aligned to 4k from Frame Buffer (local
-	 * physical) for shared ASD <-> Driver
-	 */
-	ret = amdgpu_bo_create_kernel(psp->adev, PSP_ASD_SHARED_MEM_SIZE,
-				      PAGE_SIZE, AMDGPU_GEM_DOMAIN_VRAM,
-				      &psp->asd_shared_bo,
-				      &psp->asd_shared_mc_addr,
-				      &psp->asd_shared_buf);
-
-	return ret;
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = 0;
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = 0;
+	cmd->cmd.cmd_load_ta.cmd_buf_len = 0;
 }
 
 static int psp_asd_load(struct psp_context *psp)
@@ -368,11 +350,49 @@ static int psp_asd_load(struct psp_context *psp)
 	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
 	memcpy(psp->fw_pri_buf, psp->asd_start_addr, psp->asd_ucode_size);
 
-	psp_prep_asd_cmd_buf(cmd, psp->fw_pri_mc_addr, psp->asd_shared_mc_addr,
-			     psp->asd_ucode_size, PSP_ASD_SHARED_MEM_SIZE);
+	psp_prep_asd_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
+				  psp->asd_ucode_size);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd,
+				 psp->fence_buf_mc_addr);
+	if (!ret) {
+		psp->asd_context.asd_initialized = true;
+		psp->asd_context.session_id = cmd->resp.session_id;
+	}
+
+	kfree(cmd);
+
+	return ret;
+}
+
+static void psp_prep_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+				       uint32_t session_id)
+{
+	cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA;
+	cmd->cmd.cmd_unload_ta.session_id = session_id;
+}
+
+static int psp_asd_unload(struct psp_context *psp)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	if (amdgpu_sriov_vf(psp->adev))
+		return 0;
+
+	if (!psp->asd_context.asd_initialized)
+		return 0;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_ta_unload_cmd_buf(cmd, psp->asd_context.session_id);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 				 psp->fence_buf_mc_addr);
+	if (!ret)
+		psp->asd_context.asd_initialized = false;
 
 	kfree(cmd);
 
@@ -407,18 +427,20 @@ int psp_reg_program(struct psp_context *psp, enum psp_reg_prog_id reg,
 	return ret;
 }
 
-static void psp_prep_xgmi_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					  uint64_t xgmi_ta_mc, uint64_t xgmi_mc_shared,
-					  uint32_t xgmi_ta_size, uint32_t shared_size)
+static void psp_prep_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+				     uint64_t ta_bin_mc,
+				     uint32_t ta_bin_size,
+				     uint64_t ta_shared_mc,
+				     uint32_t ta_shared_size)
 {
-        cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
-        cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(xgmi_ta_mc);
-        cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(xgmi_ta_mc);
-        cmd->cmd.cmd_load_ta.app_len = xgmi_ta_size;
+	cmd->cmd_id 				= GFX_CMD_ID_LOAD_TA;
+	cmd->cmd.cmd_load_ta.app_phy_addr_lo 	= lower_32_bits(ta_bin_mc);
+	cmd->cmd.cmd_load_ta.app_phy_addr_hi 	= upper_32_bits(ta_bin_mc);
+	cmd->cmd.cmd_load_ta.app_len 		= ta_bin_size;
 
-        cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(xgmi_mc_shared);
-        cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(xgmi_mc_shared);
-        cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(ta_shared_mc);
+	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(ta_shared_mc);
+	cmd->cmd.cmd_load_ta.cmd_buf_len 	 = ta_shared_size;
 }
 
 static int psp_xgmi_init_shared_buf(struct psp_context *psp)
@@ -438,6 +460,36 @@ static int psp_xgmi_init_shared_buf(struct psp_context *psp)
 	return ret;
 }
 
+static void psp_prep_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
+				       uint32_t ta_cmd_id,
+				       uint32_t session_id)
+{
+	cmd->cmd_id 				= GFX_CMD_ID_INVOKE_CMD;
+	cmd->cmd.cmd_invoke_cmd.session_id 	= session_id;
+	cmd->cmd.cmd_invoke_cmd.ta_cmd_id 	= ta_cmd_id;
+}
+
+int psp_ta_invoke(struct psp_context *psp,
+		  uint32_t ta_cmd_id,
+		  uint32_t session_id)
+{
+	int ret;
+	struct psp_gfx_cmd_resp *cmd;
+
+	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
+	if (!cmd)
+		return -ENOMEM;
+
+	psp_prep_ta_invoke_cmd_buf(cmd, ta_cmd_id, session_id);
+
+	ret = psp_cmd_submit_buf(psp, NULL, cmd,
+				 psp->fence_buf_mc_addr);
+
+	kfree(cmd);
+
+	return ret;
+}
+
 static int psp_xgmi_load(struct psp_context *psp)
 {
 	int ret;
@@ -446,8 +498,6 @@ static int psp_xgmi_load(struct psp_context *psp)
 	/*
 	 * TODO: bypass the loading in sriov for now
 	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
 
 	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
 	if (!cmd)
@@ -456,9 +506,11 @@ static int psp_xgmi_load(struct psp_context *psp)
 	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
 	memcpy(psp->fw_pri_buf, psp->ta_xgmi_start_addr, psp->ta_xgmi_ucode_size);
 
-	psp_prep_xgmi_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
-				      psp->xgmi_context.xgmi_shared_mc_addr,
-				      psp->ta_xgmi_ucode_size, PSP_XGMI_SHARED_MEM_SIZE);
+	psp_prep_ta_load_cmd_buf(cmd,
+				 psp->fw_pri_mc_addr,
+				 psp->ta_xgmi_ucode_size,
+				 psp->xgmi_context.xgmi_shared_mc_addr,
+				 PSP_XGMI_SHARED_MEM_SIZE);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 				 psp->fence_buf_mc_addr);
@@ -473,29 +525,25 @@ static int psp_xgmi_load(struct psp_context *psp)
 	return ret;
 }
 
-static void psp_prep_xgmi_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					    uint32_t xgmi_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA;
-	cmd->cmd.cmd_unload_ta.session_id = xgmi_session_id;
-}
-
 static int psp_xgmi_unload(struct psp_context *psp)
 {
 	int ret;
 	struct psp_gfx_cmd_resp *cmd;
+	struct amdgpu_device *adev = psp->adev;
+
+	/* XGMI TA unload currently is not supported on Arcturus */
+	if (adev->asic_type == CHIP_ARCTURUS)
+		return 0;
 
 	/*
 	 * TODO: bypass the unloading in sriov for now
 	 */
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
 
 	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
 	if (!cmd)
 		return -ENOMEM;
 
-	psp_prep_xgmi_ta_unload_cmd_buf(cmd, psp->xgmi_context.session_id);
+	psp_prep_ta_unload_cmd_buf(cmd, psp->xgmi_context.session_id);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 				 psp->fence_buf_mc_addr);
@@ -505,40 +553,9 @@ static int psp_xgmi_unload(struct psp_context *psp)
 	return ret;
 }
 
-static void psp_prep_xgmi_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					    uint32_t ta_cmd_id,
-					    uint32_t xgmi_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
-	cmd->cmd.cmd_invoke_cmd.session_id = xgmi_session_id;
-	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
-	/* Note: cmd_invoke_cmd.buf is not used for now */
-}
-
 int psp_xgmi_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
-	/*
-	 * TODO: bypass the loading in sriov for now
-	*/
-	if (amdgpu_sriov_vf(psp->adev))
-		return 0;
-
-	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
-	if (!cmd)
-		return -ENOMEM;
-
-	psp_prep_xgmi_ta_invoke_cmd_buf(cmd, ta_cmd_id,
-					psp->xgmi_context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-				 psp->fence_buf_mc_addr);
-
-	kfree(cmd);
-
-        return ret;
+	return psp_ta_invoke(psp, ta_cmd_id, psp->xgmi_context.session_id);
 }
 
 static int psp_xgmi_terminate(struct psp_context *psp)
@@ -594,20 +611,6 @@ static int psp_xgmi_initialize(struct psp_context *psp)
 }
 
 // ras begin
-static void psp_prep_ras_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-		uint64_t ras_ta_mc, uint64_t ras_mc_shared,
-		uint32_t ras_ta_size, uint32_t shared_size)
-{
-	cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
-	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(ras_ta_mc);
-	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(ras_ta_mc);
-	cmd->cmd.cmd_load_ta.app_len = ras_ta_size;
-
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(ras_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(ras_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
-}
-
 static int psp_ras_init_shared_buf(struct psp_context *psp)
 {
 	int ret;
@@ -643,15 +646,17 @@ static int psp_ras_load(struct psp_context *psp)
 	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
 	memcpy(psp->fw_pri_buf, psp->ta_ras_start_addr, psp->ta_ras_ucode_size);
 
-	psp_prep_ras_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
-			psp->ras.ras_shared_mc_addr,
-			psp->ta_ras_ucode_size, PSP_RAS_SHARED_MEM_SIZE);
+	psp_prep_ta_load_cmd_buf(cmd,
+				 psp->fw_pri_mc_addr,
+				 psp->ta_ras_ucode_size,
+				 psp->ras.ras_shared_mc_addr,
+				 PSP_RAS_SHARED_MEM_SIZE);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 			psp->fence_buf_mc_addr);
 
 	if (!ret) {
-		psp->ras.ras_initialized = 1;
+		psp->ras.ras_initialized = true;
 		psp->ras.session_id = cmd->resp.session_id;
 	}
 
@@ -660,13 +665,6 @@ static int psp_ras_load(struct psp_context *psp)
 	return ret;
 }
 
-static void psp_prep_ras_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-						uint32_t ras_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA;
-	cmd->cmd.cmd_unload_ta.session_id = ras_session_id;
-}
-
 static int psp_ras_unload(struct psp_context *psp)
 {
 	int ret;
@@ -682,7 +680,7 @@ static int psp_ras_unload(struct psp_context *psp)
 	if (!cmd)
 		return -ENOMEM;
 
-	psp_prep_ras_ta_unload_cmd_buf(cmd, psp->ras.session_id);
+	psp_prep_ta_unload_cmd_buf(cmd, psp->ras.session_id);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd,
 			psp->fence_buf_mc_addr);
@@ -692,40 +690,15 @@ static int psp_ras_unload(struct psp_context *psp)
 	return ret;
 }
 
-static void psp_prep_ras_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-		uint32_t ta_cmd_id,
-		uint32_t ras_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
-	cmd->cmd.cmd_invoke_cmd.session_id = ras_session_id;
-	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
-	/* Note: cmd_invoke_cmd.buf is not used for now */
-}
-
 int psp_ras_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
 	/*
 	 * TODO: bypass the loading in sriov for now
 	 */
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
-	if (!cmd)
-		return -ENOMEM;
-
-	psp_prep_ras_ta_invoke_cmd_buf(cmd, ta_cmd_id,
-			psp->ras.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd,
-			psp->fence_buf_mc_addr);
-
-	kfree(cmd);
-
-	return ret;
+	return psp_ta_invoke(psp, ta_cmd_id, psp->ras.session_id);
 }
 
 int psp_ras_enable_features(struct psp_context *psp,
@@ -771,7 +744,7 @@ static int psp_ras_terminate(struct psp_context *psp)
 	if (ret)
 		return ret;
 
-	psp->ras.ras_initialized = 0;
+	psp->ras.ras_initialized = false;
 
 	/* free ras shared memory */
 	amdgpu_bo_free_kernel(&psp->ras.ras_shared_bo,
@@ -812,24 +785,6 @@ static int psp_ras_initialize(struct psp_context *psp)
 // ras end
 
 // HDCP start
-static void psp_prep_hdcp_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					  uint64_t hdcp_ta_mc,
-					  uint64_t hdcp_mc_shared,
-					  uint32_t hdcp_ta_size,
-					  uint32_t shared_size)
-{
-	cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
-	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(hdcp_ta_mc);
-	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(hdcp_ta_mc);
-	cmd->cmd.cmd_load_ta.app_len = hdcp_ta_size;
-
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo =
-		lower_32_bits(hdcp_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi =
-		upper_32_bits(hdcp_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
-}
-
 static int psp_hdcp_init_shared_buf(struct psp_context *psp)
 {
 	int ret;
@@ -866,15 +821,16 @@ static int psp_hdcp_load(struct psp_context *psp)
 	memcpy(psp->fw_pri_buf, psp->ta_hdcp_start_addr,
 	       psp->ta_hdcp_ucode_size);
 
-	psp_prep_hdcp_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
-				      psp->hdcp_context.hdcp_shared_mc_addr,
-				      psp->ta_hdcp_ucode_size,
-				      PSP_HDCP_SHARED_MEM_SIZE);
+	psp_prep_ta_load_cmd_buf(cmd,
+				 psp->fw_pri_mc_addr,
+				 psp->ta_hdcp_ucode_size,
+				 psp->hdcp_context.hdcp_shared_mc_addr,
+				 PSP_HDCP_SHARED_MEM_SIZE);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
 
 	if (!ret) {
-		psp->hdcp_context.hdcp_initialized = 1;
+		psp->hdcp_context.hdcp_initialized = true;
 		psp->hdcp_context.session_id = cmd->resp.session_id;
 	}
 
@@ -910,12 +866,6 @@ static int psp_hdcp_initialize(struct psp_context *psp)
 
 	return 0;
 }
-static void psp_prep_hdcp_ta_unload_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					    uint32_t hdcp_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_UNLOAD_TA;
-	cmd->cmd.cmd_unload_ta.session_id = hdcp_session_id;
-}
 
 static int psp_hdcp_unload(struct psp_context *psp)
 {
@@ -932,7 +882,7 @@ static int psp_hdcp_unload(struct psp_context *psp)
 	if (!cmd)
 		return -ENOMEM;
 
-	psp_prep_hdcp_ta_unload_cmd_buf(cmd, psp->hdcp_context.session_id);
+	psp_prep_ta_unload_cmd_buf(cmd, psp->hdcp_context.session_id);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
 
@@ -941,39 +891,15 @@ static int psp_hdcp_unload(struct psp_context *psp)
 	return ret;
 }
 
-static void psp_prep_hdcp_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					    uint32_t ta_cmd_id,
-					    uint32_t hdcp_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
-	cmd->cmd.cmd_invoke_cmd.session_id = hdcp_session_id;
-	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
-	/* Note: cmd_invoke_cmd.buf is not used for now */
-}
-
 int psp_hdcp_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
 	/*
 	 * TODO: bypass the loading in sriov for now
 	 */
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
-	if (!cmd)
-		return -ENOMEM;
-
-	psp_prep_hdcp_ta_invoke_cmd_buf(cmd, ta_cmd_id,
-					psp->hdcp_context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	kfree(cmd);
-
-	return ret;
+	return psp_ta_invoke(psp, ta_cmd_id, psp->hdcp_context.session_id);
 }
 
 static int psp_hdcp_terminate(struct psp_context *psp)
@@ -993,7 +919,7 @@ static int psp_hdcp_terminate(struct psp_context *psp)
 	if (ret)
 		return ret;
 
-	psp->hdcp_context.hdcp_initialized = 0;
+	psp->hdcp_context.hdcp_initialized = false;
 
 	/* free hdcp shared memory */
 	amdgpu_bo_free_kernel(&psp->hdcp_context.hdcp_shared_bo,
@@ -1005,22 +931,6 @@ static int psp_hdcp_terminate(struct psp_context *psp)
 // HDCP end
 
 // DTM start
-static void psp_prep_dtm_ta_load_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					 uint64_t dtm_ta_mc,
-					 uint64_t dtm_mc_shared,
-					 uint32_t dtm_ta_size,
-					 uint32_t shared_size)
-{
-	cmd->cmd_id = GFX_CMD_ID_LOAD_TA;
-	cmd->cmd.cmd_load_ta.app_phy_addr_lo = lower_32_bits(dtm_ta_mc);
-	cmd->cmd.cmd_load_ta.app_phy_addr_hi = upper_32_bits(dtm_ta_mc);
-	cmd->cmd.cmd_load_ta.app_len = dtm_ta_size;
-
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_lo = lower_32_bits(dtm_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_phy_addr_hi = upper_32_bits(dtm_mc_shared);
-	cmd->cmd.cmd_load_ta.cmd_buf_len = shared_size;
-}
-
 static int psp_dtm_init_shared_buf(struct psp_context *psp)
 {
 	int ret;
@@ -1056,15 +966,16 @@ static int psp_dtm_load(struct psp_context *psp)
 	memset(psp->fw_pri_buf, 0, PSP_1_MEG);
 	memcpy(psp->fw_pri_buf, psp->ta_dtm_start_addr, psp->ta_dtm_ucode_size);
 
-	psp_prep_dtm_ta_load_cmd_buf(cmd, psp->fw_pri_mc_addr,
-				     psp->dtm_context.dtm_shared_mc_addr,
-				     psp->ta_dtm_ucode_size,
-				     PSP_DTM_SHARED_MEM_SIZE);
+	psp_prep_ta_load_cmd_buf(cmd,
+				 psp->fw_pri_mc_addr,
+				 psp->ta_dtm_ucode_size,
+				 psp->dtm_context.dtm_shared_mc_addr,
+				 PSP_DTM_SHARED_MEM_SIZE);
 
 	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
 
 	if (!ret) {
-		psp->dtm_context.dtm_initialized = 1;
+		psp->dtm_context.dtm_initialized = true;
 		psp->dtm_context.session_id = cmd->resp.session_id;
 	}
 
@@ -1102,39 +1013,15 @@ static int psp_dtm_initialize(struct psp_context *psp)
 	return 0;
 }
 
-static void psp_prep_dtm_ta_invoke_cmd_buf(struct psp_gfx_cmd_resp *cmd,
-					   uint32_t ta_cmd_id,
-					   uint32_t dtm_session_id)
-{
-	cmd->cmd_id = GFX_CMD_ID_INVOKE_CMD;
-	cmd->cmd.cmd_invoke_cmd.session_id = dtm_session_id;
-	cmd->cmd.cmd_invoke_cmd.ta_cmd_id = ta_cmd_id;
-	/* Note: cmd_invoke_cmd.buf is not used for now */
-}
-
 int psp_dtm_invoke(struct psp_context *psp, uint32_t ta_cmd_id)
 {
-	int ret;
-	struct psp_gfx_cmd_resp *cmd;
-
 	/*
 	 * TODO: bypass the loading in sriov for now
 	 */
 	if (amdgpu_sriov_vf(psp->adev))
 		return 0;
 
-	cmd = kzalloc(sizeof(struct psp_gfx_cmd_resp), GFP_KERNEL);
-	if (!cmd)
-		return -ENOMEM;
-
-	psp_prep_dtm_ta_invoke_cmd_buf(cmd, ta_cmd_id,
-				       psp->dtm_context.session_id);
-
-	ret = psp_cmd_submit_buf(psp, NULL, cmd, psp->fence_buf_mc_addr);
-
-	kfree(cmd);
-
-	return ret;
+	return psp_ta_invoke(psp, ta_cmd_id, psp->dtm_context.session_id);
 }
 
 static int psp_dtm_terminate(struct psp_context *psp)
@@ -1154,7 +1041,7 @@ static int psp_dtm_terminate(struct psp_context *psp)
 	if (ret)
 		return ret;
 
-	psp->dtm_context.dtm_initialized = 0;
+	psp->dtm_context.dtm_initialized = false;
 
 	/* free hdcp shared memory */
 	amdgpu_bo_free_kernel(&psp->dtm_context.dtm_shared_bo,
@@ -1211,45 +1098,6 @@ static int psp_hw_start(struct psp_context *psp)
 		return ret;
 	}
 
-	ret = psp_asd_init(psp);
-	if (ret) {
-		DRM_ERROR("PSP asd init failed!\n");
-		return ret;
-	}
-
-	ret = psp_asd_load(psp);
-	if (ret) {
-		DRM_ERROR("PSP load asd failed!\n");
-		return ret;
-	}
-
-	if (adev->gmc.xgmi.num_physical_nodes > 1) {
-		ret = psp_xgmi_initialize(psp);
-		/* Warning the XGMI seesion initialize failure
-		 * Instead of stop driver initialization
-		 */
-		if (ret)
-			dev_err(psp->adev->dev,
-				"XGMI: Failed to initialize XGMI session\n");
-	}
-
-	if (psp->adev->psp.ta_fw) {
-		ret = psp_ras_initialize(psp);
-		if (ret)
-			dev_err(psp->adev->dev,
-					"RAS: Failed to initialize RAS\n");
-
-		ret = psp_hdcp_initialize(psp);
-		if (ret)
-			dev_err(psp->adev->dev,
-				"HDCP: Failed to initialize HDCP\n");
-
-		ret = psp_dtm_initialize(psp);
-		if (ret)
-			dev_err(psp->adev->dev,
-				"DTM: Failed to initialize DTM\n");
-	}
-
 	return 0;
 }
 
@@ -1329,6 +1177,9 @@ static int psp_get_fw_type(struct amdgpu_firmware_info *ucode,
 	case AMDGPU_UCODE_ID_VCN:
 		*type = GFX_FW_TYPE_VCN;
 		break;
+	case AMDGPU_UCODE_ID_VCN1:
+		*type = GFX_FW_TYPE_VCN1;
+		break;
 	case AMDGPU_UCODE_ID_DMCU_ERAM:
 		*type = GFX_FW_TYPE_DMCU_ERAM;
 		break;
@@ -1341,6 +1192,9 @@ static int psp_get_fw_type(struct amdgpu_firmware_info *ucode,
 	case AMDGPU_UCODE_ID_VCN1_RAM:
 		*type = GFX_FW_TYPE_VCN1_RAM;
 		break;
+	case AMDGPU_UCODE_ID_DMCUB:
+		*type = GFX_FW_TYPE_DMUB;
+		break;
 	case AMDGPU_UCODE_ID_MAXIMUM:
 	default:
 		return -EINVAL;
@@ -1470,7 +1324,8 @@ out:
                     || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_G
 	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_CNTL
 	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_GPM_MEM
-	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM))
+	            || ucode->ucode_id == AMDGPU_UCODE_ID_RLC_RESTORE_LIST_SRM_MEM
+	            || ucode->ucode_id == AMDGPU_UCODE_ID_SMC))
 			/*skip ucode loading in SRIOV VF */
 			continue;
 
@@ -1519,16 +1374,13 @@ static int psp_load_fw(struct amdgpu_device *adev)
 	if (!psp->cmd)
 		return -ENOMEM;
 
-	/* this fw pri bo is not used under SRIOV */
-	if (!amdgpu_sriov_vf(psp->adev)) {
-		ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG,
-					      AMDGPU_GEM_DOMAIN_GTT,
-					      &psp->fw_pri_bo,
-					      &psp->fw_pri_mc_addr,
-					      &psp->fw_pri_buf);
-		if (ret)
-			goto failed;
-	}
+	ret = amdgpu_bo_create_kernel(adev, PSP_1_MEG, PSP_1_MEG,
+					AMDGPU_GEM_DOMAIN_GTT,
+					&psp->fw_pri_bo,
+					&psp->fw_pri_mc_addr,
+					&psp->fw_pri_buf);
+	if (ret)
+		goto failed;
 
 	ret = amdgpu_bo_create_kernel(adev, PSP_FENCE_BUFFER_SIZE, PAGE_SIZE,
 					AMDGPU_GEM_DOMAIN_VRAM,
@@ -1562,6 +1414,39 @@ skip_memalloc:
 	if (ret)
 		goto failed;
 
+	ret = psp_asd_load(psp);
+	if (ret) {
+		DRM_ERROR("PSP load asd failed!\n");
+		return ret;
+	}
+
+	if (adev->gmc.xgmi.num_physical_nodes > 1) {
+		ret = psp_xgmi_initialize(psp);
+		/* Warning the XGMI seesion initialize failure
+		 * Instead of stop driver initialization
+		 */
+		if (ret)
+			dev_err(psp->adev->dev,
+				"XGMI: Failed to initialize XGMI session\n");
+	}
+
+	if (psp->adev->psp.ta_fw) {
+		ret = psp_ras_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+					"RAS: Failed to initialize RAS\n");
+
+		ret = psp_hdcp_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+				"HDCP: Failed to initialize HDCP\n");
+
+		ret = psp_dtm_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+				"DTM: Failed to initialize DTM\n");
+	}
+
 	return 0;
 
 failed:
@@ -1619,6 +1504,8 @@ static int psp_hw_fini(void *handle)
 		psp_hdcp_terminate(psp);
 	}
 
+	psp_asd_unload(psp);
+
 	psp_ring_destroy(psp, PSP_RING_TYPE__KM);
 
 	pptr = amdgpu_sriov_vf(psp->adev) ? &tmr_buf : NULL;
@@ -1627,8 +1514,6 @@ static int psp_hw_fini(void *handle)
 			      &psp->fw_pri_mc_addr, &psp->fw_pri_buf);
 	amdgpu_bo_free_kernel(&psp->fence_buf_bo,
 			      &psp->fence_buf_mc_addr, &psp->fence_buf);
-	amdgpu_bo_free_kernel(&psp->asd_shared_bo, &psp->asd_shared_mc_addr,
-			      &psp->asd_shared_buf);
 	amdgpu_bo_free_kernel(&psp->cmd_buf_bo, &psp->cmd_buf_mc_addr,
 			      (void **)&psp->cmd_buf_mem);
 
@@ -1704,6 +1589,39 @@ static int psp_resume(void *handle)
 	if (ret)
 		goto failed;
 
+	ret = psp_asd_load(psp);
+	if (ret) {
+		DRM_ERROR("PSP load asd failed!\n");
+		goto failed;
+	}
+
+	if (adev->gmc.xgmi.num_physical_nodes > 1) {
+		ret = psp_xgmi_initialize(psp);
+		/* Warning the XGMI seesion initialize failure
+		 * Instead of stop driver initialization
+		 */
+		if (ret)
+			dev_err(psp->adev->dev,
+				"XGMI: Failed to initialize XGMI session\n");
+	}
+
+	if (psp->adev->psp.ta_fw) {
+		ret = psp_ras_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+					"RAS: Failed to initialize RAS\n");
+
+		ret = psp_hdcp_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+				"HDCP: Failed to initialize HDCP\n");
+
+		ret = psp_dtm_initialize(psp);
+		if (ret)
+			dev_err(psp->adev->dev,
+				"DTM: Failed to initialize DTM\n");
+	}
+
 	mutex_unlock(&adev->firmware.mutex);
 
 	return 0;
@@ -1758,6 +1676,56 @@ int psp_update_vcn_sram(struct amdgpu_device *adev, int inst_idx,
 	return psp_execute_np_fw_load(&adev->psp, &ucode);
 }
 
+int psp_ring_cmd_submit(struct psp_context *psp,
+			uint64_t cmd_buf_mc_addr,
+			uint64_t fence_mc_addr,
+			int index)
+{
+	unsigned int psp_write_ptr_reg = 0;
+	struct psp_gfx_rb_frame *write_frame;
+	struct psp_ring *ring = &psp->km_ring;
+	struct psp_gfx_rb_frame *ring_buffer_start = ring->ring_mem;
+	struct psp_gfx_rb_frame *ring_buffer_end = ring_buffer_start +
+		ring->ring_size / sizeof(struct psp_gfx_rb_frame) - 1;
+	struct amdgpu_device *adev = psp->adev;
+	uint32_t ring_size_dw = ring->ring_size / 4;
+	uint32_t rb_frame_size_dw = sizeof(struct psp_gfx_rb_frame) / 4;
+
+	/* KM (GPCOM) prepare write pointer */
+	psp_write_ptr_reg = psp_ring_get_wptr(psp);
+
+	/* Update KM RB frame pointer to new frame */
+	/* write_frame ptr increments by size of rb_frame in bytes */
+	/* psp_write_ptr_reg increments by size of rb_frame in DWORDs */
+	if ((psp_write_ptr_reg % ring_size_dw) == 0)
+		write_frame = ring_buffer_start;
+	else
+		write_frame = ring_buffer_start + (psp_write_ptr_reg / rb_frame_size_dw);
+	/* Check invalid write_frame ptr address */
+	if ((write_frame < ring_buffer_start) || (ring_buffer_end < write_frame)) {
+		DRM_ERROR("ring_buffer_start = %p; ring_buffer_end = %p; write_frame = %p\n",
+			  ring_buffer_start, ring_buffer_end, write_frame);
+		DRM_ERROR("write_frame is pointing to address out of bounds\n");
+		return -EINVAL;
+	}
+
+	/* Initialize KM RB frame */
+	memset(write_frame, 0, sizeof(struct psp_gfx_rb_frame));
+
+	/* Update KM RB frame */
+	write_frame->cmd_buf_addr_hi = upper_32_bits(cmd_buf_mc_addr);
+	write_frame->cmd_buf_addr_lo = lower_32_bits(cmd_buf_mc_addr);
+	write_frame->fence_addr_hi = upper_32_bits(fence_mc_addr);
+	write_frame->fence_addr_lo = lower_32_bits(fence_mc_addr);
+	write_frame->fence_value = index;
+	amdgpu_asic_flush_hdp(adev, NULL);
+
+	/* Update the write Pointer in DWORDs */
+	psp_write_ptr_reg = (psp_write_ptr_reg + rb_frame_size_dw) % ring_size_dw;
+	psp_ring_set_wptr(psp, psp_write_ptr_reg);
+	return 0;
+}
+
 static bool psp_check_fw_loading_status(struct amdgpu_device *adev,
 					enum AMDGPU_UCODE_ID ucode_type)
 {