[PATCH 2/2] drm/amdgpu: Add offset normalization in VCN v5.0.1

[Lijo Lazar]

VCN v5.0.1 also will need register offset normalization. Reuse the logic
from VCN v4.0.3. Also, avoid HDP flush similar to VCN v5.0.1

Signed-off-by: Lijo Lazar <lijo.lazar at amd.com>
Reviewed-by: Hawking Zhang <Hawking.Zhang at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c | 14 ++++++++------
 drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.h |  9 +++++++++
 drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c | 20 +++++++++++---------
 3 files changed, 28 insertions(+), 15 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c
index dda5ee187948..4aa622a2cbc9 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.c
@@ -31,6 +31,7 @@
 #include "soc15d.h"
 #include "soc15_hw_ip.h"
 #include "vcn_v2_0.h"
+#include "vcn_v4_0_3.h"
 #include "mmsch_v4_0_3.h"
 
 #include "vcn/vcn_4_0_3_offset.h"
@@ -1455,8 +1456,8 @@ static uint64_t vcn_v4_0_3_unified_ring_get_wptr(struct amdgpu_ring *ring)
 				    regUVD_RB_WPTR);
 }
 
-static void vcn_v4_0_3_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
-				uint32_t val, uint32_t mask)
+void vcn_v4_0_3_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
+				       uint32_t val, uint32_t mask)
 {
 	/* Use normalized offsets when required */
 	if (vcn_v4_0_3_normalizn_reqd(ring->adev))
@@ -1468,7 +1469,8 @@ static void vcn_v4_0_3_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t
 	amdgpu_ring_write(ring, val);
 }
 
-static void vcn_v4_0_3_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg, uint32_t val)
+void vcn_v4_0_3_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg,
+				   uint32_t val)
 {
 	/* Use normalized offsets when required */
 	if (vcn_v4_0_3_normalizn_reqd(ring->adev))
@@ -1479,8 +1481,8 @@ static void vcn_v4_0_3_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg
 	amdgpu_ring_write(ring, val);
 }
 
-static void vcn_v4_0_3_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
-				unsigned int vmid, uint64_t pd_addr)
+void vcn_v4_0_3_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
+				       unsigned int vmid, uint64_t pd_addr)
 {
 	struct amdgpu_vmhub *hub = &ring->adev->vmhub[ring->vm_hub];
 
@@ -1492,7 +1494,7 @@ static void vcn_v4_0_3_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
 					lower_32_bits(pd_addr), 0xffffffff);
 }
 
-static void vcn_v4_0_3_ring_emit_hdp_flush(struct amdgpu_ring *ring)
+void vcn_v4_0_3_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 {
 	/* VCN engine access for HDP flush doesn't work when RRMT is enabled.
 	 * This is a workaround to avoid any HDP flush through VCN ring.
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.h b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.h
index 0b046114373a..03572a1d0c9c 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.h
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v4_0_3.h
@@ -26,4 +26,13 @@
 
 extern const struct amdgpu_ip_block_version vcn_v4_0_3_ip_block;
 
+void vcn_v4_0_3_enc_ring_emit_reg_wait(struct amdgpu_ring *ring, uint32_t reg,
+				       uint32_t val, uint32_t mask);
+
+void vcn_v4_0_3_enc_ring_emit_wreg(struct amdgpu_ring *ring, uint32_t reg,
+				   uint32_t val);
+void vcn_v4_0_3_enc_ring_emit_vm_flush(struct amdgpu_ring *ring,
+				       unsigned int vmid, uint64_t pd_addr);
+void vcn_v4_0_3_ring_emit_hdp_flush(struct amdgpu_ring *ring);
+
 #endif /* __VCN_V4_0_3_H__ */
diff --git a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c
index 5472b4ff6578..c2f366a3c8cd 100644
--- a/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c
+++ b/drivers/gpu/drm/amd/amdgpu/vcn_v5_0_1.c
@@ -29,6 +29,7 @@
 #include "soc15d.h"
 #include "soc15_hw_ip.h"
 #include "vcn_v2_0.h"
+#include "vcn_v4_0_3.h"
 
 #include "vcn/vcn_5_0_0_offset.h"
 #include "vcn/vcn_5_0_0_sh_mask.h"
@@ -885,16 +886,17 @@ static const struct amdgpu_ring_funcs vcn_v5_0_1_unified_ring_vm_funcs = {
 	.get_rptr = vcn_v5_0_1_unified_ring_get_rptr,
 	.get_wptr = vcn_v5_0_1_unified_ring_get_wptr,
 	.set_wptr = vcn_v5_0_1_unified_ring_set_wptr,
-	.emit_frame_size =
-		SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
-		SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 4 +
-		4 + /* vcn_v2_0_enc_ring_emit_vm_flush */
-		5 + 5 + /* vcn_v2_0_enc_ring_emit_fence x2 vm fence */
-		1, /* vcn_v2_0_enc_ring_insert_end */
+	.emit_frame_size = SOC15_FLUSH_GPU_TLB_NUM_WREG * 3 +
+			   SOC15_FLUSH_GPU_TLB_NUM_REG_WAIT * 4 +
+			   4 + /* vcn_v2_0_enc_ring_emit_vm_flush */
+			   5 +
+			   5 + /* vcn_v2_0_enc_ring_emit_fence x2 vm fence */
+			   1, /* vcn_v2_0_enc_ring_insert_end */
 	.emit_ib_size = 5, /* vcn_v2_0_enc_ring_emit_ib */
 	.emit_ib = vcn_v2_0_enc_ring_emit_ib,
 	.emit_fence = vcn_v2_0_enc_ring_emit_fence,
-	.emit_vm_flush = vcn_v2_0_enc_ring_emit_vm_flush,
+	.emit_vm_flush = vcn_v4_0_3_enc_ring_emit_vm_flush,
+	.emit_hdp_flush = vcn_v4_0_3_ring_emit_hdp_flush,
 	.test_ring = amdgpu_vcn_enc_ring_test_ring,
 	.test_ib = amdgpu_vcn_unified_ring_test_ib,
 	.insert_nop = amdgpu_ring_insert_nop,
@@ -902,8 +904,8 @@ static const struct amdgpu_ring_funcs vcn_v5_0_1_unified_ring_vm_funcs = {
 	.pad_ib = amdgpu_ring_generic_pad_ib,
 	.begin_use = amdgpu_vcn_ring_begin_use,
 	.end_use = amdgpu_vcn_ring_end_use,
-	.emit_wreg = vcn_v2_0_enc_ring_emit_wreg,
-	.emit_reg_wait = vcn_v2_0_enc_ring_emit_reg_wait,
+	.emit_wreg = vcn_v4_0_3_enc_ring_emit_wreg,
+	.emit_reg_wait = vcn_v4_0_3_enc_ring_emit_reg_wait,
 	.emit_reg_write_reg_wait = amdgpu_ring_emit_reg_write_reg_wait_helper,
 };
 
-- 
2.25.1