drm/radeon: initial VCE support v4

Only VCE 2.0 support so far.

v2: squashing multiple patches into this one
v3: add IRQ support for CIK, major cleanups,
    basic code documentation
v4: remove HAINAN from chipset list

Signed-off-by: Christian König <christian.koenig@amd.com>

15 files changed, 1117 insertions, 9 deletions

diff --git a/drivers/gpu/drm/radeon/Makefile b/drivers/gpu/drm/radeon/Makefile
index 306364a1ecd..ed60caa3251 100644
--- a/drivers/gpu/drm/radeon/Makefile
+++ b/drivers/gpu/drm/radeon/Makefile
@@ -99,6 +99,12 @@ radeon-y += \
 	uvd_v3_1.o \
 	uvd_v4_2.o
 
+# add VCE block
+radeon-y += \
+	radeon_vce.o \
+	vce_v1_0.o \
+	vce_v2_0.o \
+
 radeon-$(CONFIG_COMPAT) += radeon_ioc32.o
 radeon-$(CONFIG_VGA_SWITCHEROO) += radeon_atpx_handler.o
 radeon-$(CONFIG_ACPI) += radeon_acpi.o
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
index e6419ca7cd3..be6eb4d9128 100644
--- a/drivers/gpu/drm/radeon/cik.c
+++ b/drivers/gpu/drm/radeon/cik.c
@@ -7490,6 +7490,20 @@ restart_ih:
 			/* reset addr and status */
 			WREG32_P(VM_CONTEXT1_CNTL2, 1, ~1);
 			break;
+		case 167: /* VCE */
+			DRM_DEBUG("IH: VCE int: 0x%08x\n", src_data);
+			switch (src_data) {
+			case 0:
+				radeon_fence_process(rdev, TN_RING_TYPE_VCE1_INDEX);
+				break;
+			case 1:
+				radeon_fence_process(rdev, TN_RING_TYPE_VCE2_INDEX);
+				break;
+			default:
+				DRM_ERROR("Unhandled interrupt: %d %d\n", src_id, src_data);
+				break;
+			}
+			break;
 		case 176: /* GFX RB CP_INT */
 		case 177: /* GFX IB CP_INT */
 			radeon_fence_process(rdev, RADEON_RING_TYPE_GFX_INDEX);
@@ -7789,6 +7803,22 @@ static int cik_startup(struct radeon_device *rdev)
 	if (r)
 		rdev->ring[R600_RING_TYPE_UVD_INDEX].ring_size = 0;
 
+	r = radeon_vce_resume(rdev);
+	if (!r) {
+		r = vce_v2_0_resume(rdev);
+		if (!r)
+			r = radeon_fence_driver_start_ring(rdev,
+							   TN_RING_TYPE_VCE1_INDEX);
+		if (!r)
+			r = radeon_fence_driver_start_ring(rdev,
+							   TN_RING_TYPE_VCE2_INDEX);
+	}
+	if (r) {
+		dev_err(rdev->dev, "VCE init error (%d).\n", r);
+		rdev->ring[TN_RING_TYPE_VCE1_INDEX].ring_size = 0;
+		rdev->ring[TN_RING_TYPE_VCE2_INDEX].ring_size = 0;
+	}
+
 	/* Enable IRQ */
 	if (!rdev->irq.installed) {
 		r = radeon_irq_kms_init(rdev);
@@ -7864,6 +7894,23 @@ static int cik_startup(struct radeon_device *rdev)
 			DRM_ERROR("radeon: failed initializing UVD (%d).\n", r);
 	}
 
+	r = -ENOENT;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	if (ring->ring_size)
+		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
+				     VCE_CMD_NO_OP);
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	if (ring->ring_size)
+		r = radeon_ring_init(rdev, ring, ring->ring_size, 0,
+				     VCE_CMD_NO_OP);
+
+	if (!r)
+		r = vce_v1_0_init(rdev);
+	else if (r != -ENOENT)
+		DRM_ERROR("radeon: failed initializing VCE (%d).\n", r);
+
 	r = radeon_ib_pool_init(rdev);
 	if (r) {
 		dev_err(rdev->dev, "IB initialization failed (%d).\n", r);
@@ -7934,6 +7981,7 @@ int cik_suspend(struct radeon_device *rdev)
 	cik_sdma_enable(rdev, false);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_suspend(rdev);
+	radeon_vce_suspend(rdev);
 	cik_fini_pg(rdev);
 	cik_fini_cg(rdev);
 	cik_irq_suspend(rdev);
@@ -8066,6 +8114,17 @@ int cik_init(struct radeon_device *rdev)
 		r600_ring_init(rdev, ring, 4096);
 	}
 
+	r = radeon_vce_init(rdev);
+	if (!r) {
+		ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+		ring->ring_obj = NULL;
+		r600_ring_init(rdev, ring, 4096);
+
+		ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+		ring->ring_obj = NULL;
+		r600_ring_init(rdev, ring, 4096);
+	}
+
 	rdev->ih.ring_obj = NULL;
 	r600_ih_ring_init(rdev, 64 * 1024);
 
@@ -8127,6 +8186,7 @@ void cik_fini(struct radeon_device *rdev)
 	radeon_irq_kms_fini(rdev);
 	uvd_v1_0_fini(rdev);
 	radeon_uvd_fini(rdev);
+	radeon_vce_fini(rdev);
 	cik_pcie_gart_fini(rdev);
 	r600_vram_scratch_fini(rdev);
 	radeon_gem_fini(rdev);
diff --git a/drivers/gpu/drm/radeon/cikd.h b/drivers/gpu/drm/radeon/cikd.h
index 98bae9d7b74..459ae021d91 100644
--- a/drivers/gpu/drm/radeon/cikd.h
+++ b/drivers/gpu/drm/radeon/cikd.h
@@ -2010,4 +2010,37 @@
 /* UVD CTX indirect */
 #define	UVD_CGC_MEM_CTRL				0xC0
 
+/* VCE */
+
+#define VCE_VCPU_CACHE_OFFSET0		0x20024
+#define VCE_VCPU_CACHE_SIZE0		0x20028
+#define VCE_VCPU_CACHE_OFFSET1		0x2002c
+#define VCE_VCPU_CACHE_SIZE1		0x20030
+#define VCE_VCPU_CACHE_OFFSET2		0x20034
+#define VCE_VCPU_CACHE_SIZE2		0x20038
+#define VCE_RB_RPTR2			0x20178
+#define VCE_RB_WPTR2			0x2017c
+#define VCE_RB_RPTR			0x2018c
+#define VCE_RB_WPTR			0x20190
+#define VCE_CLOCK_GATING_A		0x202f8
+#define VCE_CLOCK_GATING_B		0x202fc
+#define VCE_UENC_CLOCK_GATING		0x207bc
+#define VCE_UENC_REG_CLOCK_GATING	0x207c0
+#define VCE_SYS_INT_EN			0x21300
+#	define VCE_SYS_INT_TRAP_INTERRUPT_EN	(1 << 3)
+#define VCE_LMI_CTRL2			0x21474
+#define VCE_LMI_CTRL			0x21498
+#define VCE_LMI_VM_CTRL			0x214a0
+#define VCE_LMI_SWAP_CNTL		0x214b4
+#define VCE_LMI_SWAP_CNTL1		0x214b8
+#define VCE_LMI_CACHE_CTRL		0x214f4
+
+#define VCE_CMD_NO_OP		0x00000000
+#define VCE_CMD_END		0x00000001
+#define VCE_CMD_IB		0x00000002
+#define VCE_CMD_FENCE		0x00000003
+#define VCE_CMD_TRAP		0x00000004
+#define VCE_CMD_IB_AUTO		0x00000005
+#define VCE_CMD_SEMAPHORE	0x00000006
+
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon.h b/drivers/gpu/drm/radeon/radeon.h
index 024db37b183..a58a38942c7 100644
--- a/drivers/gpu/drm/radeon/radeon.h
+++ b/drivers/gpu/drm/radeon/radeon.h
@@ -113,19 +113,16 @@ extern int radeon_hard_reset;
 #define RADEONFB_CONN_LIMIT			4
 #define RADEON_BIOS_NUM_SCRATCH			8
 
-/* max number of rings */
-#define RADEON_NUM_RINGS			6
-
 /* fence seq are set to this number when signaled */
 #define RADEON_FENCE_SIGNALED_SEQ		0LL
 
 /* internal ring indices */
 /* r1xx+ has gfx CP ring */
-#define RADEON_RING_TYPE_GFX_INDEX	0
+#define RADEON_RING_TYPE_GFX_INDEX		0
 
 /* cayman has 2 compute CP rings */
-#define CAYMAN_RING_TYPE_CP1_INDEX	1
-#define CAYMAN_RING_TYPE_CP2_INDEX	2
+#define CAYMAN_RING_TYPE_CP1_INDEX		1
+#define CAYMAN_RING_TYPE_CP2_INDEX		2
 
 /* R600+ has an async dma ring */
 #define R600_RING_TYPE_DMA_INDEX		3
@@ -133,7 +130,14 @@ extern int radeon_hard_reset;
 #define CAYMAN_RING_TYPE_DMA1_INDEX		4
 
 /* R600+ */
-#define R600_RING_TYPE_UVD_INDEX	5
+#define R600_RING_TYPE_UVD_INDEX		5
+
+/* TN+ */
+#define TN_RING_TYPE_VCE1_INDEX			6
+#define TN_RING_TYPE_VCE2_INDEX			7
+
+/* max number of rings */
+#define RADEON_NUM_RINGS			8
 
 /* number of hw syncs before falling back on blocking */
 #define RADEON_NUM_SYNCS			4
@@ -1591,6 +1595,42 @@ int radeon_uvd_calc_upll_dividers(struct radeon_device *rdev,
 int radeon_uvd_send_upll_ctlreq(struct radeon_device *rdev,
                                 unsigned cg_upll_func_cntl);
 
+/*
+ * VCE
+ */
+#define RADEON_MAX_VCE_HANDLES	16
+#define RADEON_VCE_STACK_SIZE	(1024*1024)
+#define RADEON_VCE_HEAP_SIZE	(4*1024*1024)
+
+struct radeon_vce {
+	struct radeon_bo	*vcpu_bo;
+	void			*cpu_addr;
+	uint64_t		gpu_addr;
+	atomic_t		handles[RADEON_MAX_VCE_HANDLES];
+	struct drm_file		*filp[RADEON_MAX_VCE_HANDLES];
+};
+
+int radeon_vce_init(struct radeon_device *rdev);
+void radeon_vce_fini(struct radeon_device *rdev);
+int radeon_vce_suspend(struct radeon_device *rdev);
+int radeon_vce_resume(struct radeon_device *rdev);
+int radeon_vce_get_create_msg(struct radeon_device *rdev, int ring,
+			      uint32_t handle, struct radeon_fence **fence);
+int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
+			       uint32_t handle, struct radeon_fence **fence);
+void radeon_vce_free_handles(struct radeon_device *rdev, struct drm_file *filp);
+int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi);
+int radeon_vce_cs_parse(struct radeon_cs_parser *p);
+bool radeon_vce_semaphore_emit(struct radeon_device *rdev,
+			       struct radeon_ring *ring,
+			       struct radeon_semaphore *semaphore,
+			       bool emit_wait);
+void radeon_vce_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib);
+void radeon_vce_fence_emit(struct radeon_device *rdev,
+			   struct radeon_fence *fence);
+int radeon_vce_ring_test(struct radeon_device *rdev, struct radeon_ring *ring);
+int radeon_vce_ib_test(struct radeon_device *rdev, struct radeon_ring *ring);
+
 struct r600_audio_pin {
 	int			channels;
 	int			rate;
@@ -2186,6 +2226,7 @@ struct radeon_device {
 	struct radeon_gem		gem;
 	struct radeon_pm		pm;
 	struct radeon_uvd		uvd;
+	struct radeon_vce		vce;
 	uint32_t			bios_scratch[RADEON_BIOS_NUM_SCRATCH];
 	struct radeon_wb		wb;
 	struct radeon_dummy_page	dummy_page;
@@ -2205,6 +2246,7 @@ struct radeon_device {
 	const struct firmware *sdma_fw;	/* CIK SDMA firmware */
 	const struct firmware *smc_fw;	/* SMC firmware */
 	const struct firmware *uvd_fw;	/* UVD firmware */
+	const struct firmware *vce_fw;	/* VCE firmware */
 	struct r600_vram_scratch vram_scratch;
 	int msi_enabled; /* msi enabled */
 	struct r600_ih ih; /* r6/700 interrupt ring */
diff --git a/drivers/gpu/drm/radeon/radeon_asic.c b/drivers/gpu/drm/radeon/radeon_asic.c
index dda02bfc10a..4f059b2c05f 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.c
+++ b/drivers/gpu/drm/radeon/radeon_asic.c
@@ -1987,6 +1987,19 @@ static struct radeon_asic_ring ci_dma_ring = {
 	.set_wptr = &cik_sdma_set_wptr,
 };
 
+static struct radeon_asic_ring ci_vce_ring = {
+	.ib_execute = &radeon_vce_ib_execute,
+	.emit_fence = &radeon_vce_fence_emit,
+	.emit_semaphore = &radeon_vce_semaphore_emit,
+	.cs_parse = &radeon_vce_cs_parse,
+	.ring_test = &radeon_vce_ring_test,
+	.ib_test = &radeon_vce_ib_test,
+	.is_lockup = &radeon_ring_test_lockup,
+	.get_rptr = &vce_v1_0_get_rptr,
+	.get_wptr = &vce_v1_0_get_wptr,
+	.set_wptr = &vce_v1_0_set_wptr,
+};
+
 static struct radeon_asic ci_asic = {
 	.init = &cik_init,
 	.fini = &cik_fini,
@@ -2015,6 +2028,8 @@ static struct radeon_asic ci_asic = {
 		[R600_RING_TYPE_DMA_INDEX] = &ci_dma_ring,
 		[CAYMAN_RING_TYPE_DMA1_INDEX] = &ci_dma_ring,
 		[R600_RING_TYPE_UVD_INDEX] = &cayman_uvd_ring,
+		[TN_RING_TYPE_VCE1_INDEX] = &ci_vce_ring,
+		[TN_RING_TYPE_VCE2_INDEX] = &ci_vce_ring,
 	},
 	.irq = {
 		.set = &cik_irq_set,
@@ -2117,6 +2132,8 @@ static struct radeon_asic kv_asic = {
 		[R600_RING_TYPE_DMA_INDEX] = &ci_dma_ring,
 		[CAYMAN_RING_TYPE_DMA1_INDEX] = &ci_dma_ring,
 		[R600_RING_TYPE_UVD_INDEX] = &cayman_uvd_ring,
+		[TN_RING_TYPE_VCE1_INDEX] = &ci_vce_ring,
+		[TN_RING_TYPE_VCE2_INDEX] = &ci_vce_ring,
 	},
 	.irq = {
 		.set = &cik_irq_set,
diff --git a/drivers/gpu/drm/radeon/radeon_asic.h b/drivers/gpu/drm/radeon/radeon_asic.h
index ae637cfda78..13f87bf5254 100644
--- a/drivers/gpu/drm/radeon/radeon_asic.h
+++ b/drivers/gpu/drm/radeon/radeon_asic.h
@@ -863,4 +863,17 @@ bool uvd_v3_1_semaphore_emit(struct radeon_device *rdev,
 /* uvd v4.2 */
 int uvd_v4_2_resume(struct radeon_device *rdev);
 
+/* vce v1.0 */
+uint32_t vce_v1_0_get_rptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring);
+uint32_t vce_v1_0_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring);
+void vce_v1_0_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring);
+int vce_v1_0_init(struct radeon_device *rdev);
+int vce_v1_0_start(struct radeon_device *rdev);
+
+/* vce v2.0 */
+int vce_v2_0_resume(struct radeon_device *rdev);
+
 #endif
diff --git a/drivers/gpu/drm/radeon/radeon_cs.c b/drivers/gpu/drm/radeon/radeon_cs.c
index dfb5a1db87d..701ee7981b5 100644
--- a/drivers/gpu/drm/radeon/radeon_cs.c
+++ b/drivers/gpu/drm/radeon/radeon_cs.c
@@ -147,6 +147,10 @@ static int radeon_cs_get_ring(struct radeon_cs_parser *p, u32 ring, s32 priority
 	case RADEON_CS_RING_UVD:
 		p->ring = R600_RING_TYPE_UVD_INDEX;
 		break;
+	case RADEON_CS_RING_VCE:
+		/* TODO: only use the low priority ring for now */
+		p->ring = TN_RING_TYPE_VCE1_INDEX;
+		break;
 	}
 	return 0;
 }
diff --git a/drivers/gpu/drm/radeon/radeon_kms.c b/drivers/gpu/drm/radeon/radeon_kms.c
index 114d1672d61..0e078afc5db 100644
--- a/drivers/gpu/drm/radeon/radeon_kms.c
+++ b/drivers/gpu/drm/radeon/radeon_kms.c
@@ -610,6 +610,7 @@ void radeon_driver_preclose_kms(struct drm_device *dev,
 	if (rdev->cmask_filp == file_priv)
 		rdev->cmask_filp = NULL;
 	radeon_uvd_free_handles(rdev, file_priv);
+	radeon_vce_free_handles(rdev, file_priv);
 }
 
 /*
diff --git a/drivers/gpu/drm/radeon/radeon_ring.c b/drivers/gpu/drm/radeon/radeon_ring.c
index 15e44a7281a..d2980b03d1a 100644
--- a/drivers/gpu/drm/radeon/radeon_ring.c
+++ b/drivers/gpu/drm/radeon/radeon_ring.c
@@ -814,6 +814,8 @@ static int cayman_cp2_index = CAYMAN_RING_TYPE_CP2_INDEX;
 static int radeon_dma1_index = R600_RING_TYPE_DMA_INDEX;
 static int radeon_dma2_index = CAYMAN_RING_TYPE_DMA1_INDEX;
 static int r600_uvd_index = R600_RING_TYPE_UVD_INDEX;
+static int si_vce1_index = TN_RING_TYPE_VCE1_INDEX;
+static int si_vce2_index = TN_RING_TYPE_VCE2_INDEX;
 
 static struct drm_info_list radeon_debugfs_ring_info_list[] = {
 	{"radeon_ring_gfx", radeon_debugfs_ring_info, 0, &radeon_gfx_index},
@@ -822,6 +824,8 @@ static struct drm_info_list radeon_debugfs_ring_info_list[] = {
 	{"radeon_ring_dma1", radeon_debugfs_ring_info, 0, &radeon_dma1_index},
 	{"radeon_ring_dma2", radeon_debugfs_ring_info, 0, &radeon_dma2_index},
 	{"radeon_ring_uvd", radeon_debugfs_ring_info, 0, &r600_uvd_index},
+	{"radeon_ring_vce1", radeon_debugfs_ring_info, 0, &si_vce1_index},
+	{"radeon_ring_vce2", radeon_debugfs_ring_info, 0, &si_vce2_index},
 };
 
 static int radeon_debugfs_sa_info(struct seq_file *m, void *data)
diff --git a/drivers/gpu/drm/radeon/radeon_test.c b/drivers/gpu/drm/radeon/radeon_test.c
index 12e8099a082..3a13e0d1055 100644
--- a/drivers/gpu/drm/radeon/radeon_test.c
+++ b/drivers/gpu/drm/radeon/radeon_test.c
@@ -257,20 +257,36 @@ static int radeon_test_create_and_emit_fence(struct radeon_device *rdev,
 					     struct radeon_ring *ring,
 					     struct radeon_fence **fence)
 {
+	uint32_t handle = ring->idx ^ 0xdeafbeef;
 	int r;
 
 	if (ring->idx == R600_RING_TYPE_UVD_INDEX) {
-		r = radeon_uvd_get_create_msg(rdev, ring->idx, 1, NULL);
+		r = radeon_uvd_get_create_msg(rdev, ring->idx, handle, NULL);
 		if (r) {
 			DRM_ERROR("Failed to get dummy create msg\n");
 			return r;
 		}
 
-		r = radeon_uvd_get_destroy_msg(rdev, ring->idx, 1, fence);
+		r = radeon_uvd_get_destroy_msg(rdev, ring->idx, handle, fence);
 		if (r) {
 			DRM_ERROR("Failed to get dummy destroy msg\n");
 			return r;
 		}
+
+	} else if (ring->idx == TN_RING_TYPE_VCE1_INDEX ||
+		   ring->idx == TN_RING_TYPE_VCE2_INDEX) {
+		r = radeon_vce_get_create_msg(rdev, ring->idx, handle, NULL);
+		if (r) {
+			DRM_ERROR("Failed to get dummy create msg\n");
+			return r;
+		}
+
+		r = radeon_vce_get_destroy_msg(rdev, ring->idx, handle, fence);
+		if (r) {
+			DRM_ERROR("Failed to get dummy destroy msg\n");
+			return r;
+		}
+
 	} else {
 		r = radeon_ring_lock(rdev, ring, 64);
 		if (r) {
@@ -486,6 +502,16 @@ out_cleanup:
 		printk(KERN_WARNING "Error while testing ring sync (%d).\n", r);
 }
 
+static bool radeon_test_sync_possible(struct radeon_ring *ringA,
+				      struct radeon_ring *ringB)
+{
+	if (ringA->idx == TN_RING_TYPE_VCE2_INDEX &&
+	    ringB->idx == TN_RING_TYPE_VCE1_INDEX)
+		return false;
+
+	return true;
+}
+
 void radeon_test_syncing(struct radeon_device *rdev)
 {
 	int i, j, k;
@@ -500,6 +526,9 @@ void radeon_test_syncing(struct radeon_device *rdev)
 			if (!ringB->ready)
 				continue;
 
+			if (!radeon_test_sync_possible(ringA, ringB))
+				continue;
+
 			DRM_INFO("Testing syncing between rings %d and %d...\n", i, j);
 			radeon_test_ring_sync(rdev, ringA, ringB);
 
@@ -511,6 +540,12 @@ void radeon_test_syncing(struct radeon_device *rdev)
 				if (!ringC->ready)
 					continue;
 
+				if (!radeon_test_sync_possible(ringA, ringC))
+					continue;
+
+				if (!radeon_test_sync_possible(ringB, ringC))
+					continue;
+
 				DRM_INFO("Testing syncing between rings %d, %d and %d...\n", i, j, k);
 				radeon_test_ring_sync2(rdev, ringA, ringB, ringC);
 
diff --git a/drivers/gpu/drm/radeon/radeon_vce.c b/drivers/gpu/drm/radeon/radeon_vce.c
new file mode 100644
index 00000000000..2547d8ea347
--- /dev/null
+++ b/drivers/gpu/drm/radeon/radeon_vce.c
@@ -0,0 +1,588 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <linux/module.h>
+#include <drm/drmP.h>
+#include <drm/drm.h>
+
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "sid.h"
+
+/* Firmware Names */
+#define FIRMWARE_BONAIRE	"radeon/BONAIRE_vce.bin"
+
+MODULE_FIRMWARE(FIRMWARE_BONAIRE);
+
+/**
+ * radeon_vce_init - allocate memory, load vce firmware
+ *
+ * @rdev: radeon_device pointer
+ *
+ * First step to get VCE online, allocate memory and load the firmware
+ */
+int radeon_vce_init(struct radeon_device *rdev)
+{
+	unsigned long bo_size;
+	const char *fw_name;
+	int i, r;
+
+	switch (rdev->family) {
+	case CHIP_BONAIRE:
+	case CHIP_KAVERI:
+	case CHIP_KABINI:
+		fw_name = FIRMWARE_BONAIRE;
+		break;
+
+	default:
+		return -EINVAL;
+	}
+
+	r = request_firmware(&rdev->vce_fw, fw_name, rdev->dev);
+	if (r) {
+		dev_err(rdev->dev, "radeon_vce: Can't load firmware \"%s\"\n",
+			fw_name);
+		return r;
+	}
+
+	bo_size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size) +
+		  RADEON_VCE_STACK_SIZE + RADEON_VCE_HEAP_SIZE;
+	r = radeon_bo_create(rdev, bo_size, PAGE_SIZE, true,
+			     RADEON_GEM_DOMAIN_VRAM, NULL, &rdev->vce.vcpu_bo);
+	if (r) {
+		dev_err(rdev->dev, "(%d) failed to allocate VCE bo\n", r);
+		return r;
+	}
+
+	r = radeon_vce_resume(rdev);
+	if (r)
+		return r;
+
+	memset(rdev->vce.cpu_addr, 0, bo_size);
+	memcpy(rdev->vce.cpu_addr, rdev->vce_fw->data, rdev->vce_fw->size);
+
+	r = radeon_vce_suspend(rdev);
+	if (r)
+		return r;
+
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		atomic_set(&rdev->vce.handles[i], 0);
+		rdev->vce.filp[i] = NULL;
+        }
+
+	return 0;
+}
+
+/**
+ * radeon_vce_fini - free memory
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Last step on VCE teardown, free firmware memory
+ */
+void radeon_vce_fini(struct radeon_device *rdev)
+{
+	radeon_vce_suspend(rdev);
+	radeon_bo_unref(&rdev->vce.vcpu_bo);
+}
+
+/**
+ * radeon_vce_suspend - unpin VCE fw memory
+ *
+ * @rdev: radeon_device pointer
+ *
+ * TODO: Test VCE suspend/resume
+ */
+int radeon_vce_suspend(struct radeon_device *rdev)
+{
+	int r;
+
+	if (rdev->vce.vcpu_bo == NULL)
+		return 0;
+
+	r = radeon_bo_reserve(rdev->vce.vcpu_bo, false);
+	if (!r) {
+		radeon_bo_kunmap(rdev->vce.vcpu_bo);
+		radeon_bo_unpin(rdev->vce.vcpu_bo);
+		radeon_bo_unreserve(rdev->vce.vcpu_bo);
+	}
+	return r;
+}
+
+/**
+ * radeon_vce_resume - pin VCE fw memory
+ *
+ * @rdev: radeon_device pointer
+ *
+ * TODO: Test VCE suspend/resume
+ */
+int radeon_vce_resume(struct radeon_device *rdev)
+{
+	int r;
+
+	if (rdev->vce.vcpu_bo == NULL)
+		return -EINVAL;
+
+	r = radeon_bo_reserve(rdev->vce.vcpu_bo, false);
+	if (r) {
+		radeon_bo_unref(&rdev->vce.vcpu_bo);
+		dev_err(rdev->dev, "(%d) failed to reserve VCE bo\n", r);
+		return r;
+	}
+
+	r = radeon_bo_pin(rdev->vce.vcpu_bo, RADEON_GEM_DOMAIN_VRAM,
+			  &rdev->vce.gpu_addr);
+	if (r) {
+		radeon_bo_unreserve(rdev->vce.vcpu_bo);
+		radeon_bo_unref(&rdev->vce.vcpu_bo);
+		dev_err(rdev->dev, "(%d) VCE bo pin failed\n", r);
+		return r;
+	}
+
+	r = radeon_bo_kmap(rdev->vce.vcpu_bo, &rdev->vce.cpu_addr);
+	if (r) {
+		dev_err(rdev->dev, "(%d) VCE map failed\n", r);
+		return r;
+	}
+
+	radeon_bo_unreserve(rdev->vce.vcpu_bo);
+
+	return 0;
+}
+
+/**
+ * radeon_vce_free_handles - free still open VCE handles
+ *
+ * @rdev: radeon_device pointer
+ * @filp: drm file pointer
+ *
+ * Close all VCE handles still open by this file pointer
+ */
+void radeon_vce_free_handles(struct radeon_device *rdev, struct drm_file *filp)
+{
+	int i, r;
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		uint32_t handle = atomic_read(&rdev->vce.handles[i]);
+		if (!handle || rdev->vce.filp[i] != filp)
+			continue;
+
+		r = radeon_vce_get_destroy_msg(rdev, TN_RING_TYPE_VCE1_INDEX,
+					       handle, NULL);
+		if (r)
+			DRM_ERROR("Error destroying VCE handle (%d)!\n", r);
+
+		rdev->vce.filp[i] = NULL;
+		atomic_set(&rdev->vce.handles[i], 0);
+	}
+}
+
+/**
+ * radeon_vce_get_create_msg - generate a VCE create msg
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring we should submit the msg to
+ * @handle: VCE session handle to use
+ * @fence: optional fence to return
+ *
+ * Open up a stream for HW test
+ */
+int radeon_vce_get_create_msg(struct radeon_device *rdev, int ring,
+			      uint32_t handle, struct radeon_fence **fence)
+{
+	const unsigned ib_size_dw = 1024;
+	struct radeon_ib ib;
+	uint64_t dummy;
+	int i, r;
+
+	r = radeon_ib_get(rdev, ring, &ib, NULL, ib_size_dw * 4);
+	if (r) {
+		DRM_ERROR("radeon: failed to get ib (%d).\n", r);
+		return r;
+	}
+
+	dummy = ib.gpu_addr + 1024;
+
+	/* stitch together an VCE create msg */
+	ib.length_dw = 0;
+	ib.ptr[ib.length_dw++] = 0x0000000c; /* len */
+	ib.ptr[ib.length_dw++] = 0x00000001; /* session cmd */
+	ib.ptr[ib.length_dw++] = handle;
+
+	ib.ptr[ib.length_dw++] = 0x00000030; /* len */
+	ib.ptr[ib.length_dw++] = 0x01000001; /* create cmd */
+	ib.ptr[ib.length_dw++] = 0x00000000;
+	ib.ptr[ib.length_dw++] = 0x00000042;
+	ib.ptr[ib.length_dw++] = 0x0000000a;
+	ib.ptr[ib.length_dw++] = 0x00000001;
+	ib.ptr[ib.length_dw++] = 0x00000080;
+	ib.ptr[ib.length_dw++] = 0x00000060;
+	ib.ptr[ib.length_dw++] = 0x00000100;
+	ib.ptr[ib.length_dw++] = 0x00000100;
+	ib.ptr[ib.length_dw++] = 0x0000000c;
+	ib.ptr[ib.length_dw++] = 0x00000000;
+
+	ib.ptr[ib.length_dw++] = 0x00000014; /* len */
+	ib.ptr[ib.length_dw++] = 0x05000005; /* feedback buffer */
+	ib.ptr[ib.length_dw++] = upper_32_bits(dummy);
+	ib.ptr[ib.length_dw++] = dummy;
+	ib.ptr[ib.length_dw++] = 0x00000001;
+
+	for (i = ib.length_dw; i < ib_size_dw; ++i)
+		ib.ptr[i] = 0x0;
+
+	r = radeon_ib_schedule(rdev, &ib, NULL);
+	if (r) {
+	        DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
+	}
+
+	if (fence)
+		*fence = radeon_fence_ref(ib.fence);
+
+	radeon_ib_free(rdev, &ib);
+
+	return r;
+}
+
+/**
+ * radeon_vce_get_destroy_msg - generate a VCE destroy msg
+ *
+ * @rdev: radeon_device pointer
+ * @ring: ring we should submit the msg to
+ * @handle: VCE session handle to use
+ * @fence: optional fence to return
+ *
+ * Close up a stream for HW test or if userspace failed to do so
+ */
+int radeon_vce_get_destroy_msg(struct radeon_device *rdev, int ring,
+			       uint32_t handle, struct radeon_fence **fence)
+{
+	const unsigned ib_size_dw = 1024;
+	struct radeon_ib ib;
+	uint64_t dummy;
+	int i, r;
+
+	r = radeon_ib_get(rdev, ring, &ib, NULL, ib_size_dw * 4);
+	if (r) {
+		DRM_ERROR("radeon: failed to get ib (%d).\n", r);
+		return r;
+	}
+
+	dummy = ib.gpu_addr + 1024;
+
+	/* stitch together an VCE destroy msg */
+	ib.length_dw = 0;
+	ib.ptr[ib.length_dw++] = 0x0000000c; /* len */
+	ib.ptr[ib.length_dw++] = 0x00000001; /* session cmd */
+	ib.ptr[ib.length_dw++] = handle;
+
+	ib.ptr[ib.length_dw++] = 0x00000014; /* len */
+	ib.ptr[ib.length_dw++] = 0x05000005; /* feedback buffer */
+	ib.ptr[ib.length_dw++] = upper_32_bits(dummy);
+	ib.ptr[ib.length_dw++] = dummy;
+	ib.ptr[ib.length_dw++] = 0x00000001;
+
+	ib.ptr[ib.length_dw++] = 0x00000008; /* len */
+	ib.ptr[ib.length_dw++] = 0x02000001; /* destroy cmd */
+
+	for (i = ib.length_dw; i < ib_size_dw; ++i)
+		ib.ptr[i] = 0x0;
+
+	r = radeon_ib_schedule(rdev, &ib, NULL);
+	if (r) {
+	        DRM_ERROR("radeon: failed to schedule ib (%d).\n", r);
+	}
+
+	if (fence)
+		*fence = radeon_fence_ref(ib.fence);
+
+	radeon_ib_free(rdev, &ib);
+
+	return r;
+}
+
+/**
+ * radeon_vce_cs_reloc - command submission relocation
+ *
+ * @p: parser context
+ * @lo: address of lower dword
+ * @hi: address of higher dword
+ *
+ * Patch relocation inside command stream with real buffer address
+ */
+int radeon_vce_cs_reloc(struct radeon_cs_parser *p, int lo, int hi)
+{
+	struct radeon_cs_chunk *relocs_chunk;
+	uint64_t offset;
+	unsigned idx;
+
+	relocs_chunk = &p->chunks[p->chunk_relocs_idx];
+	offset = radeon_get_ib_value(p, lo);
+	idx = radeon_get_ib_value(p, hi);
+
+	if (idx >= relocs_chunk->length_dw) {
+		DRM_ERROR("Relocs at %d after relocations chunk end %d !\n",
+			  idx, relocs_chunk->length_dw);
+		return -EINVAL;
+	}
+
+	offset += p->relocs_ptr[(idx / 4)]->lobj.gpu_offset;
+
+        p->ib.ptr[lo] = offset & 0xFFFFFFFF;
+        p->ib.ptr[hi] = offset >> 32;
+
+	return 0;
+}
+
+/**
+ * radeon_vce_cs_parse - parse and validate the command stream
+ *
+ * @p: parser context
+ *
+ */
+int radeon_vce_cs_parse(struct radeon_cs_parser *p)
+{
+	uint32_t handle = 0;
+	bool destroy = false;
+	int i, r;
+
+	while (p->idx < p->chunks[p->chunk_ib_idx].length_dw) {
+		uint32_t len = radeon_get_ib_value(p, p->idx);
+		uint32_t cmd = radeon_get_ib_value(p, p->idx + 1);
+
+		if ((len < 8) || (len & 3)) {
+			DRM_ERROR("invalid VCE command length (%d)!\n", len);
+                	return -EINVAL;
+		}
+
+		switch (cmd) {
+		case 0x00000001: // session
+			handle = radeon_get_ib_value(p, p->idx + 2);
+			break;
+
+		case 0x00000002: // task info
+		case 0x01000001: // create
+		case 0x04000001: // config extension
+		case 0x04000002: // pic control
+		case 0x04000005: // rate control
+		case 0x04000007: // motion estimation
+		case 0x04000008: // rdo
+			break;
+
+		case 0x03000001: // encode
+			r = radeon_vce_cs_reloc(p, p->idx + 10, p->idx + 9);
+			if (r)
+				return r;
+
+			r = radeon_vce_cs_reloc(p, p->idx + 12, p->idx + 11);
+			if (r)
+				return r;
+			break;
+
+		case 0x02000001: // destroy
+			destroy = true;
+			break;
+
+		case 0x05000001: // context buffer
+		case 0x05000004: // video bitstream buffer
+		case 0x05000005: // feedback buffer
+			r = radeon_vce_cs_reloc(p, p->idx + 3, p->idx + 2);
+			if (r)
+				return r;
+			break;
+
+		default:
+			DRM_ERROR("invalid VCE command (0x%x)!\n", cmd);
+			return -EINVAL;
+		}
+
+		p->idx += len / 4;
+	}
+
+	if (destroy) {
+		/* IB contains a destroy msg, free the handle */
+		for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i)
+			atomic_cmpxchg(&p->rdev->vce.handles[i], handle, 0);
+
+		return 0;
+        }
+
+	/* create or encode, validate the handle */
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		if (atomic_read(&p->rdev->vce.handles[i]) == handle)
+			return 0;
+	}
+
+	/* handle not found try to alloc a new one */
+	for (i = 0; i < RADEON_MAX_VCE_HANDLES; ++i) {
+		if (!atomic_cmpxchg(&p->rdev->vce.handles[i], 0, handle)) {
+			p->rdev->vce.filp[i] = p->filp;
+			return 0;
+		}
+	}
+
+	DRM_ERROR("No more free VCE handles!\n");
+	return -EINVAL;
+}
+
+/**
+ * radeon_vce_semaphore_emit - emit a semaphore command
+ *
+ * @rdev: radeon_device pointer
+ * @ring: engine to use
+ * @semaphore: address of semaphore
+ * @emit_wait: true=emit wait, false=emit signal
+ *
+ */
+bool radeon_vce_semaphore_emit(struct radeon_device *rdev,
+			       struct radeon_ring *ring,
+			       struct radeon_semaphore *semaphore,
+			       bool emit_wait)
+{
+	uint64_t addr = semaphore->gpu_addr;
+
+	radeon_ring_write(ring, VCE_CMD_SEMAPHORE);
+	radeon_ring_write(ring, (addr >> 3) & 0x000FFFFF);
+	radeon_ring_write(ring, (addr >> 23) & 0x000FFFFF);
+	radeon_ring_write(ring, 0x01003000 | (emit_wait ? 1 : 0));
+	if (!emit_wait)
+		radeon_ring_write(ring, VCE_CMD_END);
+
+	return true;
+}
+
+/**
+ * radeon_vce_ib_execute - execute indirect buffer
+ *
+ * @rdev: radeon_device pointer
+ * @ib: the IB to execute
+ *
+ */
+void radeon_vce_ib_execute(struct radeon_device *rdev, struct radeon_ib *ib)
+{
+	struct radeon_ring *ring = &rdev->ring[ib->ring];
+	radeon_ring_write(ring, VCE_CMD_IB);
+	radeon_ring_write(ring, ib->gpu_addr);
+	radeon_ring_write(ring, upper_32_bits(ib->gpu_addr));
+	radeon_ring_write(ring, ib->length_dw);
+}
+
+/**
+ * radeon_vce_fence_emit - add a fence command to the ring
+ *
+ * @rdev: radeon_device pointer
+ * @fence: the fence
+ *
+ */
+void radeon_vce_fence_emit(struct radeon_device *rdev,
+			   struct radeon_fence *fence)
+{
+	struct radeon_ring *ring = &rdev->ring[fence->ring];
+	uint32_t addr = rdev->fence_drv[fence->ring].gpu_addr;
+
+	radeon_ring_write(ring, VCE_CMD_FENCE);
+	radeon_ring_write(ring, addr);
+	radeon_ring_write(ring, upper_32_bits(addr));
+	radeon_ring_write(ring, fence->seq);
+	radeon_ring_write(ring, VCE_CMD_TRAP);
+	radeon_ring_write(ring, VCE_CMD_END);
+}
+
+/**
+ * radeon_vce_ring_test - test if VCE ring is working
+ *
+ * @rdev: radeon_device pointer
+ * @ring: the engine to test on
+ *
+ */
+int radeon_vce_ring_test(struct radeon_device *rdev, struct radeon_ring *ring)
+{
+	uint32_t rptr = vce_v1_0_get_rptr(rdev, ring);
+	unsigned i;
+	int r;
+
+	r = radeon_ring_lock(rdev, ring, 16);
+	if (r) {
+		DRM_ERROR("radeon: vce failed to lock ring %d (%d).\n",
+			  ring->idx, r);
+		return r;
+	}
+	radeon_ring_write(ring, VCE_CMD_END);
+	radeon_ring_unlock_commit(rdev, ring);
+
+	for (i = 0; i < rdev->usec_timeout; i++) {
+	        if (vce_v1_0_get_rptr(rdev, ring) != rptr)
+	                break;
+	        DRM_UDELAY(1);
+	}
+
+	if (i < rdev->usec_timeout) {
+	        DRM_INFO("ring test on %d succeeded in %d usecs\n",
+	                 ring->idx, i);
+	} else {
+	        DRM_ERROR("radeon: ring %d test failed\n",
+	                  ring->idx);
+	        r = -ETIMEDOUT;
+	}
+
+	return r;
+}
+
+/**
+ * radeon_vce_ib_test - test if VCE IBs are working
+ *
+ * @rdev: radeon_device pointer
+ * @ring: the engine to test on
+ *
+ */
+int radeon_vce_ib_test(struct radeon_device *rdev, struct radeon_ring *ring)
+{
+	struct radeon_fence *fence = NULL;
+	int r;
+
+	r = radeon_vce_get_create_msg(rdev, ring->idx, 1, NULL);
+	if (r) {
+		DRM_ERROR("radeon: failed to get create msg (%d).\n", r);
+		goto error;
+	}
+
+	r = radeon_vce_get_destroy_msg(rdev, ring->idx, 1, &fence);
+	if (r) {
+		DRM_ERROR("radeon: failed to get destroy ib (%d).\n", r);
+		goto error;
+	}
+
+	r = radeon_fence_wait(fence, false);
+	if (r) {
+		DRM_ERROR("radeon: fence wait failed (%d).\n", r);
+	} else {
+	        DRM_INFO("ib test on ring %d succeeded\n", ring->idx);
+	}
+error:
+	radeon_fence_unref(&fence);
+	return r;
+}
diff --git a/drivers/gpu/drm/radeon/sid.h b/drivers/gpu/drm/radeon/sid.h
index 9239a6d2912..683532f8493 100644
--- a/drivers/gpu/drm/radeon/sid.h
+++ b/drivers/gpu/drm/radeon/sid.h
@@ -1798,4 +1798,51 @@
 #define	DMA_PACKET_CONSTANT_FILL			  0xd
 #define	DMA_PACKET_NOP					  0xf
 
+#define VCE_STATUS					0x20004
+#define VCE_VCPU_CNTL					0x20014
+#define		VCE_CLK_EN				(1 << 0)
+#define VCE_VCPU_CACHE_OFFSET0				0x20024
+#define VCE_VCPU_CACHE_SIZE0				0x20028
+#define VCE_VCPU_CACHE_OFFSET1				0x2002c
+#define VCE_VCPU_CACHE_SIZE1				0x20030
+#define VCE_VCPU_CACHE_OFFSET2				0x20034
+#define VCE_VCPU_CACHE_SIZE2				0x20038
+#define VCE_SOFT_RESET					0x20120
+#define 	VCE_ECPU_SOFT_RESET			(1 << 0)
+#define 	VCE_FME_SOFT_RESET			(1 << 2)
+#define VCE_RB_BASE_LO2					0x2016c
+#define VCE_RB_BASE_HI2					0x20170
+#define VCE_RB_SIZE2					0x20174
+#define VCE_RB_RPTR2					0x20178
+#define VCE_RB_WPTR2					0x2017c
+#define VCE_RB_BASE_LO					0x20180
+#define VCE_RB_BASE_HI					0x20184
+#define VCE_RB_SIZE					0x20188
+#define VCE_RB_RPTR					0x2018c
+#define VCE_RB_WPTR					0x20190
+#define VCE_CLOCK_GATING_A				0x202f8
+#define VCE_CLOCK_GATING_B				0x202fc
+#define VCE_UENC_CLOCK_GATING				0x205bc
+#define VCE_UENC_REG_CLOCK_GATING			0x205c0
+#define VCE_FW_REG_STATUS				0x20e10
+#	define VCE_FW_REG_STATUS_BUSY			(1 << 0)
+#	define VCE_FW_REG_STATUS_PASS			(1 << 3)
+#	define VCE_FW_REG_STATUS_DONE			(1 << 11)
+#define VCE_LMI_FW_START_KEYSEL				0x20e18
+#define VCE_LMI_FW_PERIODIC_CTRL			0x20e20
+#define VCE_LMI_CTRL2					0x20e74
+#define VCE_LMI_CTRL					0x20e98
+#define VCE_LMI_VM_CTRL					0x20ea0
+#define VCE_LMI_SWAP_CNTL				0x20eb4
+#define VCE_LMI_SWAP_CNTL1				0x20eb8
+#define VCE_LMI_CACHE_CTRL				0x20ef4
+
+#define VCE_CMD_NO_OP					0x00000000
+#define VCE_CMD_END					0x00000001
+#define VCE_CMD_IB					0x00000002
+#define VCE_CMD_FENCE					0x00000003
+#define VCE_CMD_TRAP					0x00000004
+#define VCE_CMD_IB_AUTO					0x00000005
+#define VCE_CMD_SEMAPHORE				0x00000006
+
 #endif
diff --git a/drivers/gpu/drm/radeon/vce_v1_0.c b/drivers/gpu/drm/radeon/vce_v1_0.c
new file mode 100644
index 00000000000..e0c3534356a
--- /dev/null
+++ b/drivers/gpu/drm/radeon/vce_v1_0.c
@@ -0,0 +1,187 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "sid.h"
+
+/**
+ * vce_v1_0_get_rptr - get read pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon_ring pointer
+ *
+ * Returns the current hardware read pointer
+ */
+uint32_t vce_v1_0_get_rptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	if (ring->idx == TN_RING_TYPE_VCE1_INDEX)
+		return RREG32(VCE_RB_RPTR);
+	else
+		return RREG32(VCE_RB_RPTR2);
+}
+
+/**
+ * vce_v1_0_get_wptr - get write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon_ring pointer
+ *
+ * Returns the current hardware write pointer
+ */
+uint32_t vce_v1_0_get_wptr(struct radeon_device *rdev,
+			   struct radeon_ring *ring)
+{
+	if (ring->idx == TN_RING_TYPE_VCE1_INDEX)
+		return RREG32(VCE_RB_WPTR);
+	else
+		return RREG32(VCE_RB_WPTR2);
+}
+
+/**
+ * vce_v1_0_set_wptr - set write pointer
+ *
+ * @rdev: radeon_device pointer
+ * @ring: radeon_ring pointer
+ *
+ * Commits the write pointer to the hardware
+ */
+void vce_v1_0_set_wptr(struct radeon_device *rdev,
+		       struct radeon_ring *ring)
+{
+	if (ring->idx == TN_RING_TYPE_VCE1_INDEX)
+		WREG32(VCE_RB_WPTR, ring->wptr);
+	else
+		WREG32(VCE_RB_WPTR2, ring->wptr);
+}
+
+/**
+ * vce_v1_0_start - start VCE block
+ *
+ * @rdev: radeon_device pointer
+ *
+ * Setup and start the VCE block
+ */
+int vce_v1_0_start(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int i, j, r;
+
+	/* set BUSY flag */
+	WREG32_P(VCE_STATUS, 1, ~1);
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	WREG32(VCE_RB_RPTR, ring->rptr);
+	WREG32(VCE_RB_WPTR, ring->wptr);
+	WREG32(VCE_RB_BASE_LO, ring->gpu_addr);
+	WREG32(VCE_RB_BASE_HI, upper_32_bits(ring->gpu_addr));
+	WREG32(VCE_RB_SIZE, ring->ring_size / 4);
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	WREG32(VCE_RB_RPTR2, ring->rptr);
+	WREG32(VCE_RB_WPTR2, ring->wptr);
+	WREG32(VCE_RB_BASE_LO2, ring->gpu_addr);
+	WREG32(VCE_RB_BASE_HI2, upper_32_bits(ring->gpu_addr));
+	WREG32(VCE_RB_SIZE2, ring->ring_size / 4);
+
+	WREG32_P(VCE_VCPU_CNTL, VCE_CLK_EN, ~VCE_CLK_EN);
+
+	WREG32_P(VCE_SOFT_RESET,
+		 VCE_ECPU_SOFT_RESET |
+		 VCE_FME_SOFT_RESET, ~(
+		 VCE_ECPU_SOFT_RESET |
+		 VCE_FME_SOFT_RESET));
+
+	mdelay(100);
+
+	WREG32_P(VCE_SOFT_RESET, 0, ~(
+		 VCE_ECPU_SOFT_RESET |
+		 VCE_FME_SOFT_RESET));
+
+	for (i = 0; i < 10; ++i) {
+		uint32_t status;
+		for (j = 0; j < 100; ++j) {
+			status = RREG32(VCE_STATUS);
+			if (status & 2)
+				break;
+			mdelay(10);
+		}
+		r = 0;
+		if (status & 2)
+			break;
+
+		DRM_ERROR("VCE not responding, trying to reset the ECPU!!!\n");
+		WREG32_P(VCE_SOFT_RESET, VCE_ECPU_SOFT_RESET, ~VCE_ECPU_SOFT_RESET);
+		mdelay(10);
+		WREG32_P(VCE_SOFT_RESET, 0, ~VCE_ECPU_SOFT_RESET);
+		mdelay(10);
+		r = -1;
+	}
+
+	/* clear BUSY flag */
+	WREG32_P(VCE_STATUS, 0, ~1);
+
+	if (r) {
+		DRM_ERROR("VCE not responding, giving up!!!\n");
+		return r;
+	}
+
+	return 0;
+}
+
+int vce_v1_0_init(struct radeon_device *rdev)
+{
+	struct radeon_ring *ring;
+	int r;
+
+	r = vce_v1_0_start(rdev);
+	if (r)
+		return r;
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE1_INDEX];
+	ring->ready = true;
+	r = radeon_ring_test(rdev, TN_RING_TYPE_VCE1_INDEX, ring);
+	if (r) {
+		ring->ready = false;
+		return r;
+	}
+
+	ring = &rdev->ring[TN_RING_TYPE_VCE2_INDEX];
+	ring->ready = true;
+	r = radeon_ring_test(rdev, TN_RING_TYPE_VCE2_INDEX, ring);
+	if (r) {
+		ring->ready = false;
+		return r;
+	}
+
+	DRM_INFO("VCE initialized successfully.\n");
+
+	return 0;
+}
diff --git a/drivers/gpu/drm/radeon/vce_v2_0.c b/drivers/gpu/drm/radeon/vce_v2_0.c
new file mode 100644
index 00000000000..4911d1b00e3
--- /dev/null
+++ b/drivers/gpu/drm/radeon/vce_v2_0.c
@@ -0,0 +1,70 @@
+/*
+ * Copyright 2013 Advanced Micro Devices, Inc.
+ * All Rights Reserved.
+ *
+ * Permission is hereby granted, free of charge, to any person obtaining a
+ * copy of this software and associated documentation files (the
+ * "Software"), to deal in the Software without restriction, including
+ * without limitation the rights to use, copy, modify, merge, publish,
+ * distribute, sub license, and/or sell copies of the Software, and to
+ * permit persons to whom the Software is furnished to do so, subject to
+ * the following conditions:
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+ * IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+ * FITNESS FOR A PARTICULAR PURPOSE AND NON-INFRINGEMENT. IN NO EVENT SHALL
+ * THE COPYRIGHT HOLDERS, AUTHORS AND/OR ITS SUPPLIERS BE LIABLE FOR ANY CLAIM,
+ * DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR
+ * OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE
+ * USE OR OTHER DEALINGS IN THE SOFTWARE.
+ *
+ * The above copyright notice and this permission notice (including the
+ * next paragraph) shall be included in all copies or substantial portions
+ * of the Software.
+ *
+ * Authors: Christian König <christian.koenig@amd.com>
+ */
+
+#include <linux/firmware.h>
+#include <drm/drmP.h>
+#include "radeon.h"
+#include "radeon_asic.h"
+#include "cikd.h"
+
+int vce_v2_0_resume(struct radeon_device *rdev)
+{
+	uint64_t addr = rdev->vce.gpu_addr;
+	uint32_t size;
+
+	WREG32_P(VCE_CLOCK_GATING_A, 0, ~(1 << 16));
+	WREG32_P(VCE_UENC_CLOCK_GATING, 0x1FF000, ~0xFF9FF000);
+	WREG32_P(VCE_UENC_REG_CLOCK_GATING, 0x3F, ~0x3F);
+	WREG32(VCE_CLOCK_GATING_B, 0xf7);
+
+	WREG32(VCE_LMI_CTRL, 0x00398000);
+	WREG32_P(VCE_LMI_CACHE_CTRL, 0x0, ~0x1);
+	WREG32(VCE_LMI_SWAP_CNTL, 0);
+	WREG32(VCE_LMI_SWAP_CNTL1, 0);
+	WREG32(VCE_LMI_VM_CTRL, 0);
+
+	size = RADEON_GPU_PAGE_ALIGN(rdev->vce_fw->size);
+	WREG32(VCE_VCPU_CACHE_OFFSET0, addr & 0x7fffffff);
+	WREG32(VCE_VCPU_CACHE_SIZE0, size);
+
+	addr += size;
+	size = RADEON_VCE_STACK_SIZE;
+	WREG32(VCE_VCPU_CACHE_OFFSET1, addr & 0x7fffffff);
+	WREG32(VCE_VCPU_CACHE_SIZE1, size);
+
+	addr += size;
+	size = RADEON_VCE_HEAP_SIZE;
+	WREG32(VCE_VCPU_CACHE_OFFSET2, addr & 0x7fffffff);
+	WREG32(VCE_VCPU_CACHE_SIZE2, size);
+
+	WREG32_P(VCE_LMI_CTRL2, 0x0, ~0x100);
+
+	WREG32_P(VCE_SYS_INT_EN, VCE_SYS_INT_TRAP_INTERRUPT_EN,
+		 ~VCE_SYS_INT_TRAP_INTERRUPT_EN);
+
+	return 0;
+}
diff --git a/include/uapi/drm/radeon_drm.h b/include/uapi/drm/radeon_drm.h
index d9ea3a73afe..6493ca55b46 100644
--- a/include/uapi/drm/radeon_drm.h
+++ b/include/uapi/drm/radeon_drm.h
@@ -919,6 +919,7 @@ struct drm_radeon_gem_va {
 #define RADEON_CS_RING_COMPUTE      1
 #define RADEON_CS_RING_DMA          2
 #define RADEON_CS_RING_UVD          3
+#define RADEON_CS_RING_VCE          4
 /* The third dword of RADEON_CHUNK_ID_FLAGS is a sint32 that sets the priority */
 /* 0 = normal, + = higher priority, - = lower priority */