[PATCH 020/102] drm/amdgpu: support hdp flush for more sdma instances

[Alex Deucher]

From: Le Ma <le.ma at amd.com>

The bit RSVD_ENG0 to RSVD_ENG5 in GPU_HDP_FLUSH_REQ/GPU_HDP_FLUSH_DONE
can be leveraged for sdma instance 2~7 to poll register/memory.

Signed-off-by: Le Ma <le.ma at amd.com>
Acked-by: Snow Zhang < Snow.Zhang at amd.com>
Signed-off-by: Alex Deucher <alexander.deucher at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h    |  6 ++++++
 drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c | 17 +++++++++++++++++
 drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c |  5 +----
 3 files changed, 24 insertions(+), 4 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index 8575ed6713c3..e9c386cec2ad 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -651,6 +651,12 @@ struct nbio_hdp_flush_reg {
 	u32 ref_and_mask_cp9;
 	u32 ref_and_mask_sdma0;
 	u32 ref_and_mask_sdma1;
+	u32 ref_and_mask_sdma2;
+	u32 ref_and_mask_sdma3;
+	u32 ref_and_mask_sdma4;
+	u32 ref_and_mask_sdma5;
+	u32 ref_and_mask_sdma6;
+	u32 ref_and_mask_sdma7;
 };
 
 struct amdgpu_mmio_remap {
diff --git a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
index bfaaa327ae3c..2e1098c5c32b 100644
--- a/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
+++ b/drivers/gpu/drm/amd/amdgpu/nbio_v7_4.c
@@ -31,6 +31,17 @@
 
 #define smnNBIF_MGCG_CTRL_LCLK	0x1013a21c
 
+/*
+ * These are nbio v7_4_1 registers mask. Temporarily define these here since
+ * nbio v7_4_1 header is incomplete.
+ */
+#define GPU_HDP_FLUSH_DONE__RSVD_ENG0_MASK	0x00001000L
+#define GPU_HDP_FLUSH_DONE__RSVD_ENG1_MASK	0x00002000L
+#define GPU_HDP_FLUSH_DONE__RSVD_ENG2_MASK	0x00004000L
+#define GPU_HDP_FLUSH_DONE__RSVD_ENG3_MASK	0x00008000L
+#define GPU_HDP_FLUSH_DONE__RSVD_ENG4_MASK	0x00010000L
+#define GPU_HDP_FLUSH_DONE__RSVD_ENG5_MASK	0x00020000L
+
 static void nbio_v7_4_remap_hdp_registers(struct amdgpu_device *adev)
 {
 	WREG32_SOC15(NBIO, 0, mmREMAP_HDP_MEM_FLUSH_CNTL,
@@ -220,6 +231,12 @@ static const struct nbio_hdp_flush_reg nbio_v7_4_hdp_flush_reg = {
 	.ref_and_mask_cp9 = GPU_HDP_FLUSH_DONE__CP9_MASK,
 	.ref_and_mask_sdma0 = GPU_HDP_FLUSH_DONE__SDMA0_MASK,
 	.ref_and_mask_sdma1 = GPU_HDP_FLUSH_DONE__SDMA1_MASK,
+	.ref_and_mask_sdma2 = GPU_HDP_FLUSH_DONE__RSVD_ENG0_MASK,
+	.ref_and_mask_sdma3 = GPU_HDP_FLUSH_DONE__RSVD_ENG1_MASK,
+	.ref_and_mask_sdma4 = GPU_HDP_FLUSH_DONE__RSVD_ENG2_MASK,
+	.ref_and_mask_sdma5 = GPU_HDP_FLUSH_DONE__RSVD_ENG3_MASK,
+	.ref_and_mask_sdma6 = GPU_HDP_FLUSH_DONE__RSVD_ENG4_MASK,
+	.ref_and_mask_sdma7 = GPU_HDP_FLUSH_DONE__RSVD_ENG5_MASK,
 };
 
 static void nbio_v7_4_detect_hw_virt(struct amdgpu_device *adev)
diff --git a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
index 1dce974e0c58..fc01502baf00 100644
--- a/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/sdma_v4_0.c
@@ -627,10 +627,7 @@ static void sdma_v4_0_ring_emit_hdp_flush(struct amdgpu_ring *ring)
 	u32 ref_and_mask = 0;
 	const struct nbio_hdp_flush_reg *nbio_hf_reg = adev->nbio_funcs->hdp_flush_reg;
 
-	if (ring->me == 0)
-		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0;
-	else
-		ref_and_mask = nbio_hf_reg->ref_and_mask_sdma1;
+	ref_and_mask = nbio_hf_reg->ref_and_mask_sdma0 << ring->me;
 
 	sdma_v4_0_wait_reg_mem(ring, 0, 1,
 			       adev->nbio_funcs->get_hdp_flush_done_offset(adev),
-- 
2.20.1