gallium/radeon: use radeon_emit

Mostly generated using a sed-script, with manual fix-up for multi-line
statements.

Reviewed-by: Marek Olšák <marek.olsak@amd.com>

9 files changed, 135 insertions, 136 deletions

diff --git a/src/gallium/drivers/r600/evergreen_hw_context.c b/src/gallium/drivers/r600/evergreen_hw_context.c
index cd073191cd..f456696970 100644
--- a/src/gallium/drivers/r600/evergreen_hw_context.c
+++ b/src/gallium/drivers/r600/evergreen_hw_context.c
@@ -68,11 +68,11 @@ void evergreen_dma_copy_buffer(struct r600_context *rctx,
 				      RADEON_PRIO_SDMA_BUFFER);
 		radeon_add_to_buffer_list(&rctx->b, &rctx->b.dma, rdst, RADEON_USAGE_WRITE,
 				      RADEON_PRIO_SDMA_BUFFER);
-		cs->buf[cs->cdw++] = DMA_PACKET(DMA_PACKET_COPY, sub_cmd, csize);
-		cs->buf[cs->cdw++] = dst_offset & 0xffffffff;
-		cs->buf[cs->cdw++] = src_offset & 0xffffffff;
-		cs->buf[cs->cdw++] = (dst_offset >> 32UL) & 0xff;
-		cs->buf[cs->cdw++] = (src_offset >> 32UL) & 0xff;
+		radeon_emit(cs, DMA_PACKET(DMA_PACKET_COPY, sub_cmd, csize));
+		radeon_emit(cs, dst_offset & 0xffffffff);
+		radeon_emit(cs, src_offset & 0xffffffff);
+		radeon_emit(cs, (dst_offset >> 32UL) & 0xff);
+		radeon_emit(cs, (src_offset >> 32UL) & 0xff);
 		dst_offset += csize << shift;
 		src_offset += csize << shift;
 		size -= csize;
diff --git a/src/gallium/drivers/r600/evergreen_state.c b/src/gallium/drivers/r600/evergreen_state.c
index 23dd475bf9..b9451963ad 100644
--- a/src/gallium/drivers/r600/evergreen_state.c
+++ b/src/gallium/drivers/r600/evergreen_state.c
@@ -1736,8 +1736,8 @@ static void evergreen_emit_db_state(struct r600_context *rctx, struct r600_atom
 		radeon_set_context_reg(cs, R_028014_DB_HTILE_DATA_BASE, a->rsurf->db_htile_data_base);
 		reloc_idx = radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx, rtex->htile_buffer,
 						  RADEON_USAGE_READWRITE, RADEON_PRIO_HTILE);
-		cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
-		cs->buf[cs->cdw++] = reloc_idx;
+		radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+		radeon_emit(cs, reloc_idx);
 	} else {
 		radeon_set_context_reg(cs, R_028ABC_DB_HTILE_SURFACE, 0);
 		radeon_set_context_reg(cs, R_028AC8_DB_PRELOAD_CONTROL, 0);
@@ -3455,17 +3455,17 @@ static void evergreen_dma_copy_tile(struct r600_context *rctx,
 				      RADEON_USAGE_READ, RADEON_PRIO_SDMA_TEXTURE);
 		radeon_add_to_buffer_list(&rctx->b, &rctx->b.dma, &rdst->resource,
 				      RADEON_USAGE_WRITE, RADEON_PRIO_SDMA_TEXTURE);
-		cs->buf[cs->cdw++] = DMA_PACKET(DMA_PACKET_COPY, sub_cmd, size);
-		cs->buf[cs->cdw++] = base >> 8;
-		cs->buf[cs->cdw++] = (detile << 31) | (array_mode << 27) |
-					(lbpp << 24) | (bank_h << 21) |
-					(bank_w << 18) | (mt_aspect << 16);
-		cs->buf[cs->cdw++] = (pitch_tile_max << 0) | ((height - 1) << 16);
-		cs->buf[cs->cdw++] = (slice_tile_max << 0);
-		cs->buf[cs->cdw++] = (x << 0) | (z << 18);
-		cs->buf[cs->cdw++] = (y << 0) | (tile_split << 21) | (nbanks << 25) | (non_disp_tiling << 28);
-		cs->buf[cs->cdw++] = addr & 0xfffffffc;
-		cs->buf[cs->cdw++] = (addr >> 32UL) & 0xff;
+		radeon_emit(cs, DMA_PACKET(DMA_PACKET_COPY, sub_cmd, size));
+		radeon_emit(cs, base >> 8);
+		radeon_emit(cs, (detile << 31) | (array_mode << 27) |
+				(lbpp << 24) | (bank_h << 21) |
+				(bank_w << 18) | (mt_aspect << 16));
+		radeon_emit(cs, (pitch_tile_max << 0) | ((height - 1) << 16));
+		radeon_emit(cs, (slice_tile_max << 0));
+		radeon_emit(cs, (x << 0) | (z << 18));
+		radeon_emit(cs, (y << 0) | (tile_split << 21) | (nbanks << 25) | (non_disp_tiling << 28));
+		radeon_emit(cs, addr & 0xfffffffc);
+		radeon_emit(cs, (addr >> 32UL) & 0xff);
 		copy_height -= cheight;
 		addr += cheight * pitch;
 		y += cheight;
diff --git a/src/gallium/drivers/r600/r600_hw_context.c b/src/gallium/drivers/r600/r600_hw_context.c
index 857da7f370..6534d8bc4a 100644
--- a/src/gallium/drivers/r600/r600_hw_context.c
+++ b/src/gallium/drivers/r600/r600_hw_context.c
@@ -115,20 +115,20 @@ void r600_flush_emit(struct r600_context *rctx)
 	}
 
 	if (rctx->b.flags & R600_CONTEXT_PS_PARTIAL_FLUSH) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_EVENT_WRITE, 0, 0);
-		cs->buf[cs->cdw++] = EVENT_TYPE(EVENT_TYPE_PS_PARTIAL_FLUSH) | EVENT_INDEX(4);
+		radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+		radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_PS_PARTIAL_FLUSH) | EVENT_INDEX(4));
 	}
 
 	if (rctx->b.chip_class >= R700 &&
 	    (rctx->b.flags & R600_CONTEXT_FLUSH_AND_INV_CB_META)) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_EVENT_WRITE, 0, 0);
-		cs->buf[cs->cdw++] = EVENT_TYPE(EVENT_TYPE_FLUSH_AND_INV_CB_META) | EVENT_INDEX(0);
+		radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+		radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_FLUSH_AND_INV_CB_META) | EVENT_INDEX(0));
 	}
 
 	if (rctx->b.chip_class >= R700 &&
 	    (rctx->b.flags & R600_CONTEXT_FLUSH_AND_INV_DB_META)) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_EVENT_WRITE, 0, 0);
-		cs->buf[cs->cdw++] = EVENT_TYPE(EVENT_TYPE_FLUSH_AND_INV_DB_META) | EVENT_INDEX(0);
+		radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+		radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_FLUSH_AND_INV_DB_META) | EVENT_INDEX(0));
 
 		/* Set FULL_CACHE_ENA for DB META flushes on r7xx and later.
 		 *
@@ -141,8 +141,8 @@ void r600_flush_emit(struct r600_context *rctx)
 
 	if (rctx->b.flags & R600_CONTEXT_FLUSH_AND_INV ||
 	    (rctx->b.chip_class == R600 && rctx->b.flags & R600_CONTEXT_STREAMOUT_FLUSH)) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_EVENT_WRITE, 0, 0);
-		cs->buf[cs->cdw++] = EVENT_TYPE(EVENT_TYPE_CACHE_FLUSH_AND_INV_EVENT) | EVENT_INDEX(0);
+		radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+		radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_CACHE_FLUSH_AND_INV_EVENT) | EVENT_INDEX(0));
 	}
 
 	if (rctx->b.flags & R600_CONTEXT_INV_CONST_CACHE) {
@@ -215,11 +215,11 @@ void r600_flush_emit(struct r600_context *rctx)
 	}
 
 	if (cp_coher_cntl) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_SURFACE_SYNC, 3, 0);
-		cs->buf[cs->cdw++] = cp_coher_cntl;   /* CP_COHER_CNTL */
-		cs->buf[cs->cdw++] = 0xffffffff;      /* CP_COHER_SIZE */
-		cs->buf[cs->cdw++] = 0;               /* CP_COHER_BASE */
-		cs->buf[cs->cdw++] = 0x0000000A;      /* POLL_INTERVAL */
+		radeon_emit(cs, PKT3(PKT3_SURFACE_SYNC, 3, 0));
+		radeon_emit(cs, cp_coher_cntl);   /* CP_COHER_CNTL */
+		radeon_emit(cs, 0xffffffff);      /* CP_COHER_SIZE */
+		radeon_emit(cs, 0);               /* CP_COHER_BASE */
+		radeon_emit(cs, 0x0000000A);      /* POLL_INTERVAL */
 	}
 
 	if (rctx->b.flags & R600_CONTEXT_START_PIPELINE_STATS) {
@@ -475,11 +475,11 @@ void r600_dma_copy_buffer(struct r600_context *rctx,
 				      RADEON_PRIO_SDMA_BUFFER);
 		radeon_add_to_buffer_list(&rctx->b, &rctx->b.dma, rdst, RADEON_USAGE_WRITE,
 				      RADEON_PRIO_SDMA_BUFFER);
-		cs->buf[cs->cdw++] = DMA_PACKET(DMA_PACKET_COPY, 0, 0, csize);
-		cs->buf[cs->cdw++] = dst_offset & 0xfffffffc;
-		cs->buf[cs->cdw++] = src_offset & 0xfffffffc;
-		cs->buf[cs->cdw++] = (dst_offset >> 32UL) & 0xff;
-		cs->buf[cs->cdw++] = (src_offset >> 32UL) & 0xff;
+		radeon_emit(cs, DMA_PACKET(DMA_PACKET_COPY, 0, 0, csize));
+		radeon_emit(cs, dst_offset & 0xfffffffc);
+		radeon_emit(cs, src_offset & 0xfffffffc);
+		radeon_emit(cs, (dst_offset >> 32UL) & 0xff);
+		radeon_emit(cs, (src_offset >> 32UL) & 0xff);
 		dst_offset += csize << 2;
 		src_offset += csize << 2;
 		size -= csize;
diff --git a/src/gallium/drivers/r600/r600_pipe.h b/src/gallium/drivers/r600/r600_pipe.h
index b8d0cf1b0e..cdb8e8211b 100644
--- a/src/gallium/drivers/r600/r600_pipe.h
+++ b/src/gallium/drivers/r600/r600_pipe.h
@@ -881,8 +881,8 @@ static inline void radeon_set_ctl_const_seq(struct radeon_winsys_cs *cs, unsigne
 {
 	assert(reg >= R600_CTL_CONST_OFFSET);
 	assert(cs->cdw+2+num <= cs->max_dw);
-	cs->buf[cs->cdw++] = PKT3(PKT3_SET_CTL_CONST, num, 0);
-	cs->buf[cs->cdw++] = (reg - R600_CTL_CONST_OFFSET) >> 2;
+	radeon_emit(cs, PKT3(PKT3_SET_CTL_CONST, num, 0));
+	radeon_emit(cs, (reg - R600_CTL_CONST_OFFSET) >> 2);
 }
 
 static inline void radeon_compute_set_context_reg(struct radeon_winsys_cs *cs, unsigned reg, unsigned value)
diff --git a/src/gallium/drivers/r600/r600_state.c b/src/gallium/drivers/r600/r600_state.c
index ab0cf5cb69..9774bdd046 100644
--- a/src/gallium/drivers/r600/r600_state.c
+++ b/src/gallium/drivers/r600/r600_state.c
@@ -1568,8 +1568,8 @@ static void r600_emit_db_state(struct r600_context *rctx, struct r600_atom *atom
 		radeon_set_context_reg(cs, R_028014_DB_HTILE_DATA_BASE, a->rsurf->db_htile_data_base);
 		reloc_idx = radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx, rtex->htile_buffer,
 						  RADEON_USAGE_READWRITE, RADEON_PRIO_HTILE);
-		cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
-		cs->buf[cs->cdw++] = reloc_idx;
+		radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+		radeon_emit(cs, reloc_idx);
 	} else {
 		radeon_set_context_reg(cs, R_028D24_DB_HTILE_SURFACE, 0);
 	}
@@ -2928,15 +2928,15 @@ static boolean r600_dma_copy_tile(struct r600_context *rctx,
 				      RADEON_PRIO_SDMA_TEXTURE);
 		radeon_add_to_buffer_list(&rctx->b, &rctx->b.dma, &rdst->resource, RADEON_USAGE_WRITE,
 				      RADEON_PRIO_SDMA_TEXTURE);
-		cs->buf[cs->cdw++] = DMA_PACKET(DMA_PACKET_COPY, 1, 0, size);
-		cs->buf[cs->cdw++] = base >> 8;
-		cs->buf[cs->cdw++] = (detile << 31) | (array_mode << 27) |
-					(lbpp << 24) | ((height - 1) << 10) |
-					pitch_tile_max;
-		cs->buf[cs->cdw++] = (slice_tile_max << 12) | (z << 0);
-		cs->buf[cs->cdw++] = (x << 3) | (y << 17);
-		cs->buf[cs->cdw++] = addr & 0xfffffffc;
-		cs->buf[cs->cdw++] = (addr >> 32UL) & 0xff;
+		radeon_emit(cs, DMA_PACKET(DMA_PACKET_COPY, 1, 0, size));
+		radeon_emit(cs, base >> 8);
+		radeon_emit(cs, (detile << 31) | (array_mode << 27) |
+				(lbpp << 24) | ((height - 1) << 10) |
+				pitch_tile_max);
+		radeon_emit(cs, (slice_tile_max << 12) | (z << 0));
+		radeon_emit(cs, (x << 3) | (y << 17));
+		radeon_emit(cs, addr & 0xfffffffc);
+		radeon_emit(cs, (addr >> 32UL) & 0xff);
 		copy_height -= cheight;
 		addr += cheight * pitch;
 		y += cheight;
diff --git a/src/gallium/drivers/r600/r600_state_common.c b/src/gallium/drivers/r600/r600_state_common.c
index b75f61b306..eed46b0922 100644
--- a/src/gallium/drivers/r600/r600_state_common.c
+++ b/src/gallium/drivers/r600/r600_state_common.c
@@ -1871,8 +1871,8 @@ static void r600_draw_vbo(struct pipe_context *ctx, const struct pipe_draw_info
 
 	/* Draw packets. */
 	if (!info.indirect) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_NUM_INSTANCES, 0, 0);
-		cs->buf[cs->cdw++] = info.instance_count;
+		radeon_emit(cs, PKT3(PKT3_NUM_INSTANCES, 0, 0));
+		radeon_emit(cs, info.instance_count);
 	}
 
 	if (unlikely(info.indirect)) {
@@ -1883,66 +1883,66 @@ static void r600_draw_vbo(struct pipe_context *ctx, const struct pipe_draw_info
 		rctx->vgt_state.last_draw_was_indirect = true;
 		rctx->last_start_instance = -1;
 
-		cs->buf[cs->cdw++] = PKT3(EG_PKT3_SET_BASE, 2, 0);
-		cs->buf[cs->cdw++] = EG_DRAW_INDEX_INDIRECT_PATCH_TABLE_BASE;
-		cs->buf[cs->cdw++] = va;
-		cs->buf[cs->cdw++] = (va >> 32UL) & 0xFF;
+		radeon_emit(cs, PKT3(EG_PKT3_SET_BASE, 2, 0));
+		radeon_emit(cs, EG_DRAW_INDEX_INDIRECT_PATCH_TABLE_BASE);
+		radeon_emit(cs, va);
+		radeon_emit(cs, (va >> 32UL) & 0xFF);
 
-		cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
-		cs->buf[cs->cdw++] = radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
-							   (struct r600_resource*)info.indirect,
-							   RADEON_USAGE_READ,
-                                                           RADEON_PRIO_DRAW_INDIRECT);
+		radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+		radeon_emit(cs, radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
+							  (struct r600_resource*)info.indirect,
+							  RADEON_USAGE_READ,
+                                                          RADEON_PRIO_DRAW_INDIRECT));
 	}
 
 	if (info.indexed) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_INDEX_TYPE, 0, 0);
-		cs->buf[cs->cdw++] = ib.index_size == 4 ?
-					(VGT_INDEX_32 | (R600_BIG_ENDIAN ? VGT_DMA_SWAP_32_BIT : 0)) :
-					(VGT_INDEX_16 | (R600_BIG_ENDIAN ? VGT_DMA_SWAP_16_BIT : 0));
+		radeon_emit(cs, PKT3(PKT3_INDEX_TYPE, 0, 0));
+		radeon_emit(cs, ib.index_size == 4 ?
+				(VGT_INDEX_32 | (R600_BIG_ENDIAN ? VGT_DMA_SWAP_32_BIT : 0)) :
+				(VGT_INDEX_16 | (R600_BIG_ENDIAN ? VGT_DMA_SWAP_16_BIT : 0)));
 
 		if (ib.user_buffer) {
 			unsigned size_bytes = info.count*ib.index_size;
 			unsigned size_dw = align(size_bytes, 4) / 4;
-			cs->buf[cs->cdw++] = PKT3(PKT3_DRAW_INDEX_IMMD, 1 + size_dw, render_cond_bit);
-			cs->buf[cs->cdw++] = info.count;
-			cs->buf[cs->cdw++] = V_0287F0_DI_SRC_SEL_IMMEDIATE;
+			radeon_emit(cs, PKT3(PKT3_DRAW_INDEX_IMMD, 1 + size_dw, render_cond_bit));
+			radeon_emit(cs, info.count);
+			radeon_emit(cs, V_0287F0_DI_SRC_SEL_IMMEDIATE);
 			memcpy(cs->buf+cs->cdw, ib.user_buffer, size_bytes);
 			cs->cdw += size_dw;
 		} else {
 			uint64_t va = r600_resource(ib.buffer)->gpu_address + ib.offset;
 
 			if (likely(!info.indirect)) {
-				cs->buf[cs->cdw++] = PKT3(PKT3_DRAW_INDEX, 3, render_cond_bit);
-				cs->buf[cs->cdw++] = va;
-				cs->buf[cs->cdw++] = (va >> 32UL) & 0xFF;
-				cs->buf[cs->cdw++] = info.count;
-				cs->buf[cs->cdw++] = V_0287F0_DI_SRC_SEL_DMA;
-				cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
-				cs->buf[cs->cdw++] = radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
-									   (struct r600_resource*)ib.buffer,
-									   RADEON_USAGE_READ,
-                                                                           RADEON_PRIO_INDEX_BUFFER);
+				radeon_emit(cs, PKT3(PKT3_DRAW_INDEX, 3, render_cond_bit));
+				radeon_emit(cs, va);
+				radeon_emit(cs, (va >> 32UL) & 0xFF);
+				radeon_emit(cs, info.count);
+				radeon_emit(cs, V_0287F0_DI_SRC_SEL_DMA);
+				radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+				radeon_emit(cs, radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
+									  (struct r600_resource*)ib.buffer,
+									  RADEON_USAGE_READ,
+                                                                          RADEON_PRIO_INDEX_BUFFER));
 			}
 			else {
 				uint32_t max_size = (ib.buffer->width0 - ib.offset) / ib.index_size;
 
-				cs->buf[cs->cdw++] = PKT3(EG_PKT3_INDEX_BASE, 1, 0);
-				cs->buf[cs->cdw++] = va;
-				cs->buf[cs->cdw++] = (va >> 32UL) & 0xFF;
+				radeon_emit(cs, PKT3(EG_PKT3_INDEX_BASE, 1, 0));
+				radeon_emit(cs, va);
+				radeon_emit(cs, (va >> 32UL) & 0xFF);
 
-				cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
-				cs->buf[cs->cdw++] = radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
-									   (struct r600_resource*)ib.buffer,
-									   RADEON_USAGE_READ,
-                                                                           RADEON_PRIO_INDEX_BUFFER);
+				radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+				radeon_emit(cs, radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
+									  (struct r600_resource*)ib.buffer,
+									  RADEON_USAGE_READ,
+                                                                          RADEON_PRIO_INDEX_BUFFER));
 
-				cs->buf[cs->cdw++] = PKT3(EG_PKT3_INDEX_BUFFER_SIZE, 0, 0);
-				cs->buf[cs->cdw++] = max_size;
+				radeon_emit(cs, PKT3(EG_PKT3_INDEX_BUFFER_SIZE, 0, 0));
+				radeon_emit(cs, max_size);
 
-				cs->buf[cs->cdw++] = PKT3(EG_PKT3_DRAW_INDEX_INDIRECT, 1, render_cond_bit);
-				cs->buf[cs->cdw++] = info.indirect_offset;
-				cs->buf[cs->cdw++] = V_0287F0_DI_SRC_SEL_DMA;
+				radeon_emit(cs, PKT3(EG_PKT3_DRAW_INDEX_INDIRECT, 1, render_cond_bit));
+				radeon_emit(cs, info.indirect_offset);
+				radeon_emit(cs, V_0287F0_DI_SRC_SEL_DMA);
 			}
 		}
 	} else {
@@ -1952,29 +1952,29 @@ static void r600_draw_vbo(struct pipe_context *ctx, const struct pipe_draw_info
 
 			radeon_set_context_reg(cs, R_028B30_VGT_STRMOUT_DRAW_OPAQUE_VERTEX_STRIDE, t->stride_in_dw);
 
-			cs->buf[cs->cdw++] = PKT3(PKT3_COPY_DW, 4, 0);
-			cs->buf[cs->cdw++] = COPY_DW_SRC_IS_MEM | COPY_DW_DST_IS_REG;
-			cs->buf[cs->cdw++] = va & 0xFFFFFFFFUL;     /* src address lo */
-			cs->buf[cs->cdw++] = (va >> 32UL) & 0xFFUL; /* src address hi */
-			cs->buf[cs->cdw++] = R_028B2C_VGT_STRMOUT_DRAW_OPAQUE_BUFFER_FILLED_SIZE >> 2; /* dst register */
-			cs->buf[cs->cdw++] = 0; /* unused */
+			radeon_emit(cs, PKT3(PKT3_COPY_DW, 4, 0));
+			radeon_emit(cs, COPY_DW_SRC_IS_MEM | COPY_DW_DST_IS_REG);
+			radeon_emit(cs, va & 0xFFFFFFFFUL);     /* src address lo */
+			radeon_emit(cs, (va >> 32UL) & 0xFFUL); /* src address hi */
+			radeon_emit(cs, R_028B2C_VGT_STRMOUT_DRAW_OPAQUE_BUFFER_FILLED_SIZE >> 2); /* dst register */
+			radeon_emit(cs, 0); /* unused */
 
-			cs->buf[cs->cdw++] = PKT3(PKT3_NOP, 0, 0);
-			cs->buf[cs->cdw++] = radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
-								   t->buf_filled_size, RADEON_USAGE_READ,
-								   RADEON_PRIO_SO_FILLED_SIZE);
+			radeon_emit(cs, PKT3(PKT3_NOP, 0, 0));
+			radeon_emit(cs, radeon_add_to_buffer_list(&rctx->b, &rctx->b.gfx,
+								  t->buf_filled_size, RADEON_USAGE_READ,
+								  RADEON_PRIO_SO_FILLED_SIZE));
 		}
 
 		if (likely(!info.indirect)) {
-			cs->buf[cs->cdw++] = PKT3(PKT3_DRAW_INDEX_AUTO, 1, render_cond_bit);
-			cs->buf[cs->cdw++] = info.count;
+			radeon_emit(cs, PKT3(PKT3_DRAW_INDEX_AUTO, 1, render_cond_bit));
+			radeon_emit(cs, info.count);
 		}
 		else {
-			cs->buf[cs->cdw++] = PKT3(EG_PKT3_DRAW_INDIRECT, 1, render_cond_bit);
-			cs->buf[cs->cdw++] = info.indirect_offset;
+			radeon_emit(cs, PKT3(EG_PKT3_DRAW_INDIRECT, 1, render_cond_bit));
+			radeon_emit(cs, info.indirect_offset);
 		}
-		cs->buf[cs->cdw++] = V_0287F0_DI_SRC_SEL_AUTO_INDEX |
-					(info.count_from_stream_output ? S_0287F0_USE_OPAQUE(1) : 0);
+		radeon_emit(cs, V_0287F0_DI_SRC_SEL_AUTO_INDEX |
+				(info.count_from_stream_output ? S_0287F0_USE_OPAQUE(1) : 0));
 	}
 
 	/* SMX returns CONTEXT_DONE too early workaround */
@@ -1991,8 +1991,8 @@ static void r600_draw_vbo(struct pipe_context *ctx, const struct pipe_draw_info
 
 	/* ES ring rolling over at EOP - workaround */
 	if (rctx->b.chip_class == R600) {
-		cs->buf[cs->cdw++] = PKT3(PKT3_EVENT_WRITE, 0, 0);
-		cs->buf[cs->cdw++] = EVENT_TYPE(EVENT_TYPE_SQ_NON_EVENT);
+		radeon_emit(cs, PKT3(PKT3_EVENT_WRITE, 0, 0));
+		radeon_emit(cs, EVENT_TYPE(EVENT_TYPE_SQ_NON_EVENT));
 	}
 
 	/* Set the depth buffer as dirty. */
diff --git a/src/gallium/drivers/radeon/radeon_uvd.c b/src/gallium/drivers/radeon/radeon_uvd.c
index b775cd3408..a2d1d2d746 100644
--- a/src/gallium/drivers/radeon/radeon_uvd.c
+++ b/src/gallium/drivers/radeon/radeon_uvd.c
@@ -100,9 +100,8 @@ static void flush(struct ruvd_decoder *dec)
 /* add a new set register command to the IB */
 static void set_reg(struct ruvd_decoder *dec, unsigned reg, uint32_t val)
 {
-	uint32_t *pm4 =	dec->cs->buf;
-	pm4[dec->cs->cdw++] = RUVD_PKT0(reg >> 2, 0);
-	pm4[dec->cs->cdw++] = val;
+	radeon_emit(dec->cs, RUVD_PKT0(reg >> 2, 0));
+	radeon_emit(dec->cs, val);
 }
 
 /* send a command to the VCPU through the GPCOM registers */
diff --git a/src/gallium/drivers/radeonsi/cik_sdma.c b/src/gallium/drivers/radeonsi/cik_sdma.c
index cd7b0e6c23..d8ec2a3262 100644
--- a/src/gallium/drivers/radeonsi/cik_sdma.c
+++ b/src/gallium/drivers/radeonsi/cik_sdma.c
@@ -51,15 +51,15 @@ static void cik_sdma_do_copy_buffer(struct si_context *ctx,
 
 	for (i = 0; i < ncopy; i++) {
 		csize = MIN2(size, CIK_SDMA_COPY_MAX_SIZE);
-		cs->buf[cs->cdw++] = CIK_SDMA_PACKET(CIK_SDMA_OPCODE_COPY,
-						     CIK_SDMA_COPY_SUB_OPCODE_LINEAR,
-						     0);
-		cs->buf[cs->cdw++] = csize;
-		cs->buf[cs->cdw++] = 0; /* src/dst endian swap */
-		cs->buf[cs->cdw++] = src_offset;
-		cs->buf[cs->cdw++] = src_offset >> 32;
-		cs->buf[cs->cdw++] = dst_offset;
-		cs->buf[cs->cdw++] = dst_offset >> 32;
+		radeon_emit(cs, CIK_SDMA_PACKET(CIK_SDMA_OPCODE_COPY,
+						CIK_SDMA_COPY_SUB_OPCODE_LINEAR,
+						0));
+		radeon_emit(cs, csize);
+		radeon_emit(cs, 0); /* src/dst endian swap */
+		radeon_emit(cs, src_offset);
+		radeon_emit(cs, src_offset >> 32);
+		radeon_emit(cs, dst_offset);
+		radeon_emit(cs, dst_offset >> 32);
 		dst_offset += csize;
 		src_offset += csize;
 		size -= csize;
diff --git a/src/gallium/drivers/radeonsi/si_dma.c b/src/gallium/drivers/radeonsi/si_dma.c
index 1ba61f55dd..4291361a6d 100644
--- a/src/gallium/drivers/radeonsi/si_dma.c
+++ b/src/gallium/drivers/radeonsi/si_dma.c
@@ -68,11 +68,11 @@ static void si_dma_copy_buffer(struct si_context *ctx,
 
 	for (i = 0; i < ncopy; i++) {
 		csize = size < max_csize ? size : max_csize;
-		cs->buf[cs->cdw++] = SI_DMA_PACKET(SI_DMA_PACKET_COPY, sub_cmd, csize);
-		cs->buf[cs->cdw++] = dst_offset;
-		cs->buf[cs->cdw++] = src_offset;
-		cs->buf[cs->cdw++] = (dst_offset >> 32UL) & 0xff;
-		cs->buf[cs->cdw++] = (src_offset >> 32UL) & 0xff;
+		radeon_emit(cs, SI_DMA_PACKET(SI_DMA_PACKET_COPY, sub_cmd, csize));
+		radeon_emit(cs, dst_offset);
+		radeon_emit(cs, src_offset);
+		radeon_emit(cs, (dst_offset >> 32UL) & 0xff);
+		radeon_emit(cs, (src_offset >> 32UL) & 0xff);
 		dst_offset += csize << shift;
 		src_offset += csize << shift;
 		size -= csize;
@@ -164,17 +164,17 @@ static void si_dma_copy_tile(struct si_context *ctx,
 			cheight = (SI_DMA_COPY_MAX_SIZE_DW * 4) / pitch;
 		}
 		size = (cheight * pitch) / 4;
-		cs->buf[cs->cdw++] = SI_DMA_PACKET(SI_DMA_PACKET_COPY, sub_cmd, size);
-		cs->buf[cs->cdw++] = base >> 8;
-		cs->buf[cs->cdw++] = (detile << 31) | (array_mode << 27) |
-					(lbpp << 24) | (bank_h << 21) |
-					(bank_w << 18) | (mt_aspect << 16);
-		cs->buf[cs->cdw++] = (pitch_tile_max << 0) | ((height - 1) << 16);
-		cs->buf[cs->cdw++] = (slice_tile_max << 0) | (pipe_config << 26);
-		cs->buf[cs->cdw++] = (tiled_x << 0) | (tiled_z << 18);
-		cs->buf[cs->cdw++] = (tiled_y << 0) | (tile_split << 21) | (nbanks << 25) | (mt << 27);
-		cs->buf[cs->cdw++] = addr & 0xfffffffc;
-		cs->buf[cs->cdw++] = (addr >> 32UL) & 0xff;
+		radeon_emit(cs, SI_DMA_PACKET(SI_DMA_PACKET_COPY, sub_cmd, size));
+		radeon_emit(cs, base >> 8);
+		radeon_emit(cs, (detile << 31) | (array_mode << 27) |
+				(lbpp << 24) | (bank_h << 21) |
+				(bank_w << 18) | (mt_aspect << 16));
+		radeon_emit(cs, (pitch_tile_max << 0) | ((height - 1) << 16));
+		radeon_emit(cs, (slice_tile_max << 0) | (pipe_config << 26));
+		radeon_emit(cs, (tiled_x << 0) | (tiled_z << 18));
+		radeon_emit(cs, (tiled_y << 0) | (tile_split << 21) | (nbanks << 25) | (mt << 27));
+		radeon_emit(cs, addr & 0xfffffffc);
+		radeon_emit(cs, (addr >> 32UL) & 0xff);
 		copy_height -= cheight;
 		addr += cheight * pitch;
 		tiled_y += cheight;