[PATCH 1/5] drm/amdgpu:add switch buffer to end of CS

[Monk Liu]

sync switch buffer scheme with windows kmd for gfx v8,
Now always and only insert one switch buffer to the
end of CS.

Change-Id: Ief8539b2ad91ccb38b9adbfb54e27d8282f3a3bd
Signed-off-by: Monk Liu <Monk.Liu at amd.com>
---
 drivers/gpu/drm/amd/amdgpu/amdgpu.h    |  2 ++
 drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c |  2 ++
 drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c  | 13 +++++++------
 3 files changed, 11 insertions(+), 6 deletions(-)

diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu.h b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
index cb0098a..a935831 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu.h
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu.h
@@ -338,6 +338,7 @@ struct amdgpu_ring_funcs {
 	void (*end_use)(struct amdgpu_ring *ring);
 	void (*emit_wreg) (struct amdgpu_ring *ring, uint32_t offset, uint32_t val);
 	void (*emit_rreg) (struct amdgpu_ring *ring, uint32_t offset);
+	void (*emit_switch_buffer) (struct amdgpu_ring *ring);
 };
 
 /*
@@ -2372,6 +2373,7 @@ amdgpu_get_sdma_instance(struct amdgpu_ring *ring)
 #define amdgpu_ring_emit_hdp_invalidate(r) (r)->funcs->emit_hdp_invalidate((r))
 #define amdgpu_ring_emit_wreg(r, i, v) (r)->funcs->emit_wreg((r), (i), (v))
 #define amdgpu_ring_emit_rreg(r, i) (r)->funcs->emit_rreg((r), (i))
+#define amdgpu_ring_emit_switch_buffer(r) (r)->funcs->emit_switch_buffer((r))
 #define amdgpu_ring_pad_ib(r, ib) ((r)->funcs->pad_ib((r), (ib)))
 #define amdgpu_ring_init_cond_exec(r) (r)->funcs->init_cond_exec((r))
 #define amdgpu_ring_patch_cond_exec(r,o) (r)->funcs->patch_cond_exec((r),(o))
diff --git a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
index a31d7ef..029ee79 100644
--- a/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
+++ b/drivers/gpu/drm/amd/amdgpu/amdgpu_ib.c
@@ -210,6 +210,8 @@ int amdgpu_ib_schedule(struct amdgpu_ring *ring, unsigned num_ibs,
 		amdgpu_ring_patch_cond_exec(ring, patch_offset);
 
 	ring->current_ctx = ctx;
+	if (ring->funcs->emit_switch_buffer)
+		amdgpu_ring_emit_switch_buffer(ring);
 	amdgpu_ring_commit(ring);
 	return 0;
 }
diff --git a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
index dfa2288..26fced0 100755
--- a/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
+++ b/drivers/gpu/drm/amd/amdgpu/gfx_v8_0.c
@@ -5936,12 +5936,6 @@ static void gfx_v8_0_ring_emit_ib_gfx(struct amdgpu_ring *ring,
 {
 	u32 header, control = 0;
 
-	/* insert SWITCH_BUFFER packet before first IB in the ring frame */
-	if (ctx_switch) {
-		amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
-		amdgpu_ring_write(ring, 0);
-	}
-
 	if (ib->flags & AMDGPU_IB_FLAG_CE)
 		header = PACKET3(PACKET3_INDIRECT_BUFFER_CONST, 2);
 	else
@@ -6170,6 +6164,12 @@ static void gfx_v8_0_ring_emit_wreg_kiq(struct amdgpu_ring *ring, u32 idx, u32 v
 	amdgpu_ring_write(ring, val);
 }
 
+static void gfx_v8_ring_emit_sb(struct amdgpu_ring *ring)
+{
+	amdgpu_ring_write(ring, PACKET3(PACKET3_SWITCH_BUFFER, 0));
+	amdgpu_ring_write(ring, 0);
+}
+
 static void gfx_v8_0_set_gfx_eop_interrupt_state(struct amdgpu_device *adev,
 						 enum amdgpu_interrupt_state state)
 {
@@ -6477,6 +6477,7 @@ static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_gfx = {
 	.test_ib = gfx_v8_0_ring_test_ib,
 	.insert_nop = amdgpu_ring_insert_nop,
 	.pad_ib = amdgpu_ring_generic_pad_ib,
+	.emit_switch_buffer = gfx_v8_ring_emit_sb,
 };
 
 static const struct amdgpu_ring_funcs gfx_v8_0_ring_funcs_compute = {
-- 
1.9.1